Assignment

Learning Objectives

· Discuss the general history of automation and robots

· Discuss the applications of robots in various industries

· Differentiate between industrial and consumer applications of robots

· Identify common components of robots

· Discuss impacts of robots on future jobs

· Identify legal issues related to robotics

Chapter 2 briefly introduced robotics, an early and practical application of concepts developed in AI. In this chapter, we present a number of applications of robots in industrial as well as personal settings. Besides learning about the already deployed and emerging applications, we identify the general components of a robot. In the spirit of managerial considerations, we also discuss the impact of robotics on jobs as well as related legal issues. Some of the coverage is broad and impacts all other artificial intelligence (AI), so it may seem to overlap a bit with Chapter 14. But the focus in this chapter is on physical robots, not just software-driven applications of AI.

This chapter has the following sections:

1. 10.1 Opening Vignette: Robots Provide Emotional Support to Patients and Children 581

2. 10.2 Overview of Robotics 584

3. 10.3 History of Robotics 584

4. 10.4 Illustrative Applications of Robotics 586

5. 10.5 Components of Robots 595

6. 10.6 Various Categories of Robots 596

7. 10.7 Autonomous Cars: Robots in Motion 597

8. 10.8 Impact of Robots on Current and Future Jobs 600

9. 10.9 Legal Implications of Robots and Artificial Intelligence 603

Emotional Support to Patients and Children

As discussed in this chapter, robots have impacted industrial manufacturing and other physical activities. Now, with the research and evolution of AI, robotics can straddle the social world. For example, hospitals today make an effort to give social and emotional support to patients and their families. This support is especially sensitive when offering treatment to children. Children in a hospital are in an unfamiliar environment with medical instruments attached to them, and in many cases, doctors may recommend movement restrictions. This restriction leads to stress, anxiety, and depression in children and consequently in their family members. Hospitals try to provide childcare support specialist or companion pet therapies to reduce the trauma. These therapies prepare children and their parents for future treatment and provide them with temporary emotional support with their interactions. Due to the small number of such specialists, there is a gap between demand and supply for childcare specialists. Also, it is not possible to provide pet therapy at many centers due to the fear of allergies, dust, and bites that may cause the patient’s condition to be aggravated. To fill these gaps, the use of social robots is being explored to resolve depression and anxiety among children. A study (Jeong et al., 2015) found that the physical presence of a robot is more effective concerning emotional response as compared to a virtual machine interaction in a pediatric hospital center.

Researchers have known for a long time (e.g.,  Goris et al., 2010 ) that more than 60 percent of human communication is not verbal but rather occurs through facial expressions. Thus, a social robot has to be able to provide emotional communication like a child specialist. One popular robot that is providing such support is Huggable. With the help of AI, Huggable is equipped to understand facial expressions, temperament, gestures, and human cleverness. It is like a staff member added to the team of specialists who provide children some general emotional health assistance.

Huggable looks like a teddy bear having a ringed arrangement. A furry soft body provides a childish look to it and hence is perceived as a friend by the children. With its mechanical arms, Huggable can perform specific actions quickly. Rather than sporting high-tech devices, a Huggable robot is composed of an Android device whose microphone, speaker, and camera are in its internal sensors, and a mobile phone that acts as the central nervous system. The Android device enables the communication between the internal sensors and teleoperation interface. Its segmental arm components enable an easy replacement of sensors and hence increase its reusability. These haptic sensors along with AI enable it to process the physical touch and use it expressively.

Sensors incorporated in a Huggable transmit physical touch and pressure data to the teleoperation device or external device via an IOIO board. The Android device receives the data from the external sensors and transmits them to the motors that are attached to the body of the robot. These motors enable the movement of the robot. The capacitors are placed at various parts of the robot, known as pressure points. These pressure points enable the robot to understand the pain of a child who is unable to express it verbally but may be able to touch the robot to convey the pain. The Android device interprets the physical touch and pressure sensor data in a meaningful way and responds effectively. The Android phone enables communication between the other devices while keeping the design minimalistic. The computing power of the robot and the Android device is good enough to allow real-time communication with a child.  Figure 10.1  exhibits a schematic of the Huggable robot.

Figure 10.1 A General Schematic of a Huggable Robot.

Figure 10.1 Full Alternative Text

Huggable has been used with children undergoing treatment at Boston Children’s Hospital. Reportedly, Aurora, a 10-year-old who had leukemia, was being treated at Dana-Farber/Boston Children’s Cancer and Blood Disorder Center. According to Aurora’s parents, “There were many activities to do at the hospital but the Huggable being there is great for kids.” Beatrice, another child who visits the hospital frequently due to her chronic condition, misses her classes and friends and is unable to do anything that a typical child of her age would do. She was nervous and disliked the process of treatment, but during her interplay with the Huggable, she was more willing to take medicine as if it were the most natural activity to do. She recommended the robot to be a bit faster so that the next time she could play peek-a-boo correctly.

During these interactions with Huggable, children were seen hugging it, holding its hand, tickling it, giving it high-fives, and treating it as someone they need for support. Children were polite with it and used expressions such as “no, thank you” and “one second, please.” In the end, when bidding it goodbye, one child hugged the Huggable, and another wished to play with it longer.

Another benefit of such emotional support robots is in the prevention of infections. Patients may have contagious diseases, but the robots are sterilized after each use to prevent infection from spreading. Thus, Huggable not only provides support to children but also can be a useful tool for reducing the spread of infectious diseases.

A recently reported study by researchers at MIT’s Media Lab highlighted the differences between social robots such as Huggable and other virtual interaction technologies. A group of 54 children who were in a hospital were given three distinct social interactions: a cute regular teddy bear, a virtual persona of Huggable on a tablet, and a social robot. The bear offered a physical model but not social dialogue. A virtual version of Huggable on the tablet provided linguistic engagement, conversed with the humans in the same way, and possessed the same features as the robot but was a 3D virtual version of the Huggable robot. Both the virtual character and robot were operated by a teleoperator, and hence they perceived the interaction and responded in the same manner. Children were given one of the three interactions to play with based in groups according to age and gender. Necessary information was provided to the children by the care specialist, and the virtual character and robot were handled separately by these specialists just outside the room. IBM Watson’s tone Analyzer was used to attempt to identify five human emotions and five personality traits. Interactions with each of these three types of virtual agents by children were videotaped and analyzed by the researchers. The results of this experiment were quite interesting. These results showed that the children gazed more at the virtual character and the robot as compared to the bear. Touches between the children and the virtual agents were the highest with the Huggable robot followed by the virtual character Huggable and the bear. Also, the children took care of the Huggable robot and did not push or pull it. Interestingly, a few children responded to the virtual character on the tablet violently even when it made ouch sounds. The poor teddy bear was thrown and kicked around playfully. These results show that the children connected with the robotic Huggable more than the other two options.

Social Robot for Older Adults: Paro

Major countries in the world will soon have population rates of people aged 65 and older exceed that of the younger population by 2050. The emotional support older adults need cannot be ignored where geographical separation and the technological divide have made it difficult for them to connect with their families. Paro, a social robot, is designed to interact with humans and is marketed as a robot used for older adults at nursing homes. Paro mainly acts as pet therapy; it can also be immensely useful where pet therapy becomes inadvisable in hospitals due to the risk of infections.

Paro interprets the human touch, and it can also capture limited speech, express a limited set of vocal utterances, and move its head. Paro is not a mobile robot and resembles a seal. It was tested at two nursing homes (Broekens et al., 2009) with 23 patients. The results showed that social robots like Paro increase the social interaction. Paro not only brought smiles to patients’ faces but also some vivid, happy experiences to occupants. Even though Paro did not provide a complete response that humans do, many patients found the responses meaningful and connected emotionally to it. These robots can help break the monotonous routine of older adults and add some joy to their lives. It provides them with a feeling of being wanted and self-esteem, lowering stress and anxiety levels.

Questions for the Opening Vignette

1. What characteristics would you expect to have in a robot that provides emotional support to patients?

2. Can you think of other applications where robots such as the Huggable can play a helpful role?

3. Visit the website  https://www.universal-robots.com/case-stories/aurolab/  to learn about collaborative robots. How could such robots be useful in other settings?

What We Can Learn from This Vignette

As we have seen in various chapters throughout this book, AI is opening many interesting and unique applications. The stories about the Huggable and Paro introduce us to the idea of using robots for one of the most difficult aspects of work – to provide emotional support to patients, both children and adults. Combinations of technologies such as machine learning, voice synthesis, voice recognition, natural language processing, machine vision, automation, micromachines, and so on make it possible to combine these technologies to satisfy many needs. The applications can come entirely in virtual forms such as IBM Watson, which won the Jeopardy! game implementing industrial automation, producing self-driving cars, and even providing emotional support as noted in this opening vignette. We will see many examples of similar applications in this chapter.

Sources:  J. Broekens, M. Heerink, & H. Rosendal. (2009) . “Assistive Social Robots in Elderly Care: A Review.” Gerontechnology, 8, pp. 94–103. doi: 10.4017/gt.2009.08.02.002.00;  S. Fallon. (2015) . “A Blue Robotic Bear to Make Sick Kids Feel Less Blue.”  https://www.wired.com/2015/03/blue-robotic-bear-make-sick-kids-feel-less-blue/  (accessed August 2018). Also see the YouTube video at  https://youtu.be/UaRCCA2rRR0  (accessed August 2018);  K. Goris et al. (2010, September) . “Mechanical Design of the Huggable Robot Probo.” Robotics & Multibody Mechanics Research Group. Brussels, Belgium: Vrije Universiteit Brussel;  S. Jeong et al. (2015) . “A Social Robot to Mitigate Stress, Anxiety, and Pain in Hospital Pediatric Care.” Proceedings of the Tenth Annual ACM/IEEE International Conference on Human-Robot Interaction Extended Abstracts;  S. Jeong & D. Logan. (2018, April 21–26) . “Huggable: The Impact of Embodiment on Promoting Socio-emotional Interactions for Young Pediatric Surgeons.” MIT Media Lab, Cambridge, MA, CHI 2018, Montréal, Quebec, Canada.

10.2 Overview of Robotics

Every robotics scientist has her or his own view about the definition of robot. But a common notion of robot is a machine or a physical device or software that with the cooperation of AI can accomplish a responsibility autonomously. A robot can sense and affect the environment. Applications of robotics in our day-to-day lives have been increasing. This evolution and use of technologies are called the fourth industrial revolution. Applications of robotics in manufacturing, health, and information technology (IT) fields in the last decade have led to rapid development in changing the future of industries. Robots are moving from just performing preselected repetitive tasks (automation) and being unable to react to unforeseen circumstances (Ayres and Miller, 1981) to performing specialized tasks in healthcare, manufacturing, sports, financial services – virtually every industry. This capability of adaptation to new situations leads to autonomy, a sea change from previous generations of robots. Chapter 2 introduced a definition of robots and provided some applications in selected industries. In this chapter, we will supplement that introduction with various applications and take a slightly deeper dive into the topic.

Although our imagination of a robot may be based on the R2D2 or C3-PO from the Star Wars movies, we have experienced robots in many other ways. Factories have been using robots for a long time (see Section 10.3) for manufacturing. On the consumer side, an early application was Roomba, a robot that can clean floors on its own. Perhaps the best example of robots that we will all experience soon if not already is an autonomous (self-driving) car. Tech Republic called the self-driving car the first robot we will all learn to trust. We will dig a bit deeper into self-driving vehicles in Section 10.7. With the growth in machine learning, especially image recognition systems, applications of robots are increasing in virtually every industry. Robots can cut sausages into the right size pieces for pizza and can automatically determine that the right number and type of peperoni pieces have been placed on a pizza before it is baked. Surgeries conducted by and with the assistance of robots are growing at a rapid pace. Section 10.4 provides many illustrative applications of robots. Then Section 10.7 discusses self-driving cars as another category of robots.

Section 10.2 Review Questions

1. Define robot.

2. What is the difference between automation and autonomy?

3. Give examples of robots in use. Find recent applications online and share with the class.

10.3 History of Robotics

Wikipedia includes an interesting history of robotics. Humans have been fascinated with the idea of machines serving us for a long time. The first idea of robotics was conceptualized in 320 BC when Aristotle, a Greek philosopher, stated, “If every tool, when ordered, or even of its own accord, could do the work that befits it, then there would be no need either of apprentices for the master workers or of slaves for the lords.” In 1495, Leonardo Da Vinci drafted strategies and images for a robot that looked like a human. Between 1700 and 1900, various automatons were created, including an excellent automation structure built by Jacques De Vaucanson, who made one clockwork duck that could flap its wings, quack, and appear to eat and digest food.

Throughout the industrial revolution, robotics was triggered by the advances in steam power and electricity. As consumer demand increased, engineers strove to devise new methods to increase production by automation and create machines that can perform the tasks that were dangerous for a human to do. In 1893, “Steam Man,” a prototype for a humanoid robot, was proposed by Canadian professor George Moore. It was composed of steel and powered by a steam engine. It could walk autonomously at nearly nine miles per hour and could even pull relatively light loads. In 1898, Nikola Tesla exhibited a submarine prototype. These events led to the integration of robotics in manufacturing, space, defense, aerospace, medicine, education, and entertainment industries.

In 1913, the world’s first moving conveyor belt assembly line was started by Henry Ford. With the aid of a conveyor belt, a car could be assembled in 93 minutes. Later in 1920, the term  robot  was coined by Karel Capek in his play Rossum’s Universal Robots. Then a toy robot, Lilliput, was manufactured in Japan.

By the 1950s, innovators were creating machines that could handle dangerous, repetitive tasks for defense and industrial manufacturing. Since the robots were primarily designed for heavy-duty industries, they were required to pull, lift, move, and push the same way humans did. Thus, many robots were designed like a human arm. Examples include a spray-painting gadget for a position-controlling apparatus by W. L. V. Pollard in 1938. DeVilbiss Company acquired this robot and later became a leading supplier of the robotic arms in the United States.

In the mid-1950s, the first commercial robotic arm, Planetbot, was developed, and General Motors later used it in a manufacturing plant for the production of radiators. A total of eight Planetbots were sold. According to the company, it could perform nearly 25 movements and could be reset in minutes to perform another set of operations. However, Planetbot did not achieve the desired results due to the unusual behavior of the hydraulic fuel inside it.

George Devol and Joe Engelberger designed Unimate to automate the manufacturing of TV picture tubes. It weighed close to 4,000 pounds and was controlled by preprogrammed commands fed on a magnetic drum. Later this was used by General Motors Corporation for production to sequence and stack hot die-cast metal components. This arm with specific upgrades became one of the famous features in assembly lines. A total of 8,500 machines was sold, and half of them went to the automotive industries. Later Unimate was modified to perform spot welding, die casting, and machine tool stacking.

In the 1960s, Ralph Mosher and his team created two remotely operated robotic arms, Handyman and Man-mate. A Handyman was a two-arm electro-hydraulic robot, and the design of the Man-mate’s arm was based on the human spine. The arms gave the robots the flexibility for artifact examination procedures. The fingers were designed in a way that they could grasp objects via a single command.

New mobile robots came into the picture. The first one, Shakey, was developed in 1963. It could move freely, avoiding obstacles in its path. A radio antenna was attached to its head. It had a vision sensor atop a central processing unit. Shakey was attached to two wheels, and its two sensors could sense obstacles. Using logic-based problem solving, it could recognize the shape of objects, move them, or go around them.

The space race started by Russia’s Sputnik and embraced by the United States led to many technology advances leading to growth of robotics. In 1976, during NASA’s mission to Mars, a Viking lander was created for the atmospheric conditions of Mars. Its arms opened out and created a tube to gather samples from the Mars surface. There were some technical issues during the mission, but the scientists were able to fix them remotely.

In 1986, the first LEGO-based educational products were put on the market by Honda. In 1994, Dante II, an eight-legged walking robot built by Carnegie Mellon University, collected the volcanic gas sample from Mount Spur.

Robotics expanded exponentially as more research and money were invested. Robotic applications and research spread to Japan, Korea, and European nations. It is estimated that by 2019, there will be close to 2.6 million significant robots. Robots have applications in the fields of social support, defense, toys and entertainment, healthcare, food, and rescue. Many robots are now moving into next stages, going from deep-sea to interplanetary and extrasolar research. And as noted, self-driving cars will bring robots to the masses. We review several robot applications in the following sections.

Section 10.3 Review Questions

1. Identify some of the key milestones in the history of manufacturing that have led to the current interest in robotics.

2. How would Shakey’s capabilities compare to today’s robots?

3. How have robots helped with space missions?

10.4 Illustrative Applications of Robotics

This section highlights examples of robot applications in various industries. Each of these is presented as a mini-application case, with the discussion questions presented at the end of the section.

Changing Precision Technology

A mobile production company in China, Changing Precision Technology switched to the use of robotic arms to produce parts for mobile phones. The company previously employed 650 workers to operate the factory. Now, robots perform most of its operations, and the company has reduced its workforce to 60, decreasing the human workforce by 90 percent. In the future, the company intends to drop its employee count to about 20. With the robots in place, the company not only has achieved an increase in production of 250 percent but also cut the defect levels from 25 percent to a mere 5 percent.

Compiled from C. Forrest. (2015). “Chinese Factory Replaces 90%90% of Humans with Robots, Production Soars.” TechRepublic. https://www.techrepublic.com/article/chinese-factory-replaces-90-of-humans-with-robots-production-soars/ (accessed September 2018); J. Javelosa & K. Houser. (2017). “Production Soars for Chinese Factory Who Replaced 90%90% of Employees with Robots.” Future Society. https://futurism.com/2-production-soars-for-chinese-factory-who-replaced-90-of-employees-with-robots/ (accessed September 2018).

Adidas

Adidas is a worldwide leading sportswear manufacturer. Keeping trends, innovation, and customization in mind, Adidas has started to automate factories such as Speedfactory in Ansbach, Germany, and Atlanta, Georgia. A conventional supply chain from the raw materials to final product takes around two months, but with automation, it takes just a few days or weeks. The implementation of robotics there was different from that of other manufacturing industries because the raw materials used in shoes manufactured by Adidas are soft textile materials. Adidas is working with the company Oechsler to implement the robotics in its supply chain. Adidas uses technologies such as additive manufacturing, robotic arms, and computerized knitting. At the Speedfactory, the robot that makes a part of a sneaker attaches a scannable QR code to the part. During quality check, if any part of the product turns out to be faulty, the robot that created it is thus traceable and repaired. Adidas has optimized this process, which offers the company the option to roll out a few thousands of customized shoes in the market and see how it performs and optimize the process accordingly. In the next few years, the company plans to roll out around 1 million pairs of the custom styles annually. In the long term, this strategy supports moving from manufacturing large stocks of inventory to creating the products on demand.

Compiled from “Adidas’s High-Tech Factory Brings Production Back to Germany.” (2017, January 14). The Economist. https://www.economist.com/business/2017/01/14/adidass-high-tech-factory-brings-production-back-to-germany (accessed September 2018); D. Green. (2018). “Adidas Just Opened a Futuristic New Factory – and It Will Dramatically Change How Shoes Are Sold.” Business Insider. http://www.businessinsider.com/adidas-high-tech-speedfactory-begins-production-2018-4 (accessed September 2018).

BMW Employs Collaborative Robots

The increased use of AI and automation in industries has resulted in the development of robots. Yet, human cognitive capabilities are irreplaceable. The combination of robots and humans has been achieved using collaborative robots at a BMW manufacturing unit. By doing so, the company has maximized the efficiency of its production unit and modernized the work environment.

BMW’s Spartanburg, South Carolina, plant has employed 60 collaborative robots that work side by side with its human workforce. These robots, for example, furnish the interior of BMW car doors with sound and moisture insulation. This sealing protects the electronic equipment that is fixed on the door and the vehicle as a whole from moisture. Previously, human workers performed this intensive task of fixing the foil with the adhesive beads by using a manual roller. With the use of cobots, a robot’s arms perform this task with precision. Cobots run on low speed and stop immediately as soon as the sensors detect any obstacle in their way to maintain the safety of assembly-line workers.

At BMW’s Dingolfing factory located in Germany, a lightweight cobot is ceiling mounted in the axle transmission assembly area to pick up bevel gears. These gears can weigh up to 5.5 kilos. The cobot fits the bevel gears accurately, avoiding damage to the gear wheels.

Compiled from M. Allinson. (2017, March 4). “BMW Shows Off Its Smart Factory Technologies at Its Plants Worldwide.” BMW Press Release. Robotics and Automation. https://roboticsandautomationnews.com/2017/03/04/bmw-shows-off-its-smart-factory-technologies-at-its-plants-worldwide/11696/ (accessed September 2018); “Innovative Human-Robot Cooperation in BMW Group Production.” (2013, October 9). https://www.press.bmwgroup.com/global/article/detail/T0209722EN/innovative-human-robot-cooperation-in-bmw-group-production?language=en (accessed September 2018).

Tega

Tega is a social bot intended to provide extended support to preschoolers by engaging them via storytelling and offering help with vocabulary. Like Huggable, Tega is an Android-based robot and resembles an animation character. It has an external camera and onboard speakers and is designed to run for up to six hours before needing a recharge. Tega uses Android capabilities for expressive eyes, computation abilities, and physical movements. Children’s response is fed to the Tega as a reward signal into a reinforcement learning algorithm. Tega uses a social controller, sensor processing, and motor control for moving its body and tilting and rotating left or right.

Tega is designed not only to tell stories but also to hold a conversation about the stories. With the help of an app on a tablet, Tega interacts with a child as a peer and teammate, not as an educator. Children communicate with the tablet, and Tega provides the feedback and reactions by watching the children’s emotional states. Tega also offers help with vocabulary and understands a child’s physical and emotional responses, enabling it to build a relationship with the child. The tests have shown that Tega can positively impact a child’s interest in education, free thinking, and mental development. For more information, watch the video at https://www.youtube.com/watch?v=16in922JTsw.

Compiled from E. Ackerman. (2016). IEEE Spectrum. http://spectrum.ieee.org/automaton/robotics/home-robots/tega-mit-latest-friendly-squishable-social-robot (March 5, 2017); J. K. Westlund et al. (2016). “Tega: A Social Robot.” Video Presentation. Proceedings of the Eleventh ACM/IEEE International Conference on Human Robot Interaction; H. W. Park et al. (2017). “Growing Growth Mindset with a Social Robot Peer.” Proceedings of the Twelfth ACM/IEEE International Conference on Human Robot Interaction; Personal Robots Group. (2016). https://www.youtube.com/watch?v=sF0tRCqvyT0 (accessed September 2018); Personal Robots Group, MIT Media Lab. (2016). AAAS. https://www.eurekalert.org/pub_releases/2016-03/nsf-rlc031116.php (accessed September 2018).

San Francisco Burger Eatery

Flipping burgers is considered a low-pay, mundane task that provides many people with employment at a low salary. Such jobs are likely to disappear over time because of robots. One such implementation of robotics in the food industry is at a burger restaurant in San Francisco. The burger-making machine is not a traditional robot sporting arms and legs that can move around and work as a human. Instead, it is a complete burger prep device that can work from prepping a burger for cooking and bringing together a full meal. It blends the robotic power in bringing the right taste with the help of a Michelin-star chef’s recipes and being friendly on the pocket. The restaurant has put in place two 14-foot-long machines that can make around 120 burgers per hour. Each machine has 350 sensors, 20 computers, and close to 7,000 parts.

Buns, onions, tomatoes, pickles, seasoning, and sauces are filled in transparent tubes over a conveyor belt. Once an order is placed via a mobile device, it takes close to five minutes to prepare the order. First, air pressure pushes a burger brioche roll from the transparent tube on the conveyor belt. Different components of the robot work one after the other to prepare the order, from slicing the roll in two halves, applying butter on the bun, shredding vegetables, and dropping the sauces. Also, a light specialized grip is placed on the patty to keep it intact and to bake it per the recipe. With the use of thermal sensors and an algorithm, the cooking time and temperature of the patty are determined, and once cooked, the patty is placed on the bun by a robotic arm. Workers receive a notification via an Apple watch when there is an issue with the machine regarding a malfunction on an order or the need for refills on supplies.

Compiled from “A Robot Cooks Burgers at Startup Restaurant Creator.” (2018). TechCrunch. https://techcrunch.com/video/a-robot-cooks-burgers-at-startup-restaurant-creator/ (accessed September 2018); L. Zimberoff. (2018, June 21). “A Burger Joint Where Robots Make Your Food.” https://www.wsj.com/articles/a-burger-joint-where-robots-make-your-food-1529599213 (accessed September 2018).

Spyce

Using robots to make affordable foods is demonstrated by a fast-food restaurant operating in Boston that serves grain dishes and salad bowls. Spyce is a budget-friendly restaurant founded by MIT engineering graduates. Michael Farid created the robots that can cook. This restaurant employs few people with good pay and employs robots to do much of the fast-food work.

Orders are placed at a kiosk with touch screens. Once the order is confirmed, the mechanized systems start preparing the food. Ingredients are placed in refrigerated bins that are passed via transparent tubes and are collected using a mobile device that delivers the ingredients to the requested pot. A metal plate attached to the side of the robotic pot heats the food. A temperature of about 450 degrees Fahrenheit is maintained, and the food is tumbled for nearly two minutes and cooked. This resembles clothes being washed in a machine. Once the meal is ready, the robotic pot tilts and transfers the food to a bowl. After each cooking round, the robotic pot washes itself with a high-pressure hot water stream and then returns to its initial position, ready to cook the next meal. The customer name is also added to the bowl. The meal is then served by a human after any final changes. Spyce is also trying to put in place a robot that can cook pancakes.

Compiled from B. Coxworth. (2018, May 29). “Restaurant Keeps Its Prices Down – With a Robotic Kitchen.” New Atlas. https://newatlas.com/spyce-restaurant-robotic-kitchen/54818/ (accessed September 2018); J. Engel. (2018, May 3). “Spyce, MIT-Born Robotic Kitchen Startup, Launches Restaurant: Video.” Xconomy. https://www.xconomy.com/boston/2018/05/03/spyce-mit-born-robotic-kitchen-startup-launches-restaurant-video/ (accessed September 2018).

Mahindra & Mahindra Ltd.

As the population increases, the agricultural industry is expanding to keep up with demand. To keep increasing the food supply at a reasonable cost and to maintain quality, the Indian multinational firm Mahindra & Mahindra Ltd. is seeking to improve the process of harvesting tabletop grapes. The company is establishing a research and development center at Virginia Polytechnic Institute and University. It will work with other Mahindra centers situated in Finland, India, and Japan.

The grapes can be used for juice, wine, and tabletop grapes. The quality that must be maintained is vastly different for each of these. The ripeness and presentation of tabletop grapes differ from the other two uses; hence, quality control is critical. Deciding which grapes are ready to pick is a labor-intensive approach, and one must ensure the maturity, consistency, and quality of grapes. Making this decision visually requires expert training, which is not easily scalable. Using robotic harvesting instead of human pickers is being explored. Robots can achieve these goals using sensors that will keep the quality in view while speeding the process.

Compiled from L. Rosencrance. (2018, May 31). “Tabletop Grapes to Get Picked by Robots in India, with Help from Virginia Tech.” RoboticsBusinessReview. https://www.roboticsbusinessreview.com/agriculture/tabletop-grapes-picked-robots-india-virginia-tech/ (accessed September 2018); “Tabletop Grapes to Get Picked by Robots in India.” Agtechnews.com. http://agtechnews.com/Ag-Robotics-Technology/Tabletop-Grapes-to-Get-Picked-by-Robots-in-India.html (accessed September 2018).

Robots in the Defense Industry

For obvious reasons, the military has invested in robotic applications for a long time. Robots can replace humans in places where risk of loss of human life is too great. Robots can also reach areas where humans may not be able to go due to extreme conditions – heat, water, and so forth. Besides the recent growth of drones in military applications, several specific robots have been developed over a long time. Some are highlighted in the next sections.

MAARS

MAARS (Modular Advanced Armed Robotic System) is an upgraded version of special weapons observation reconnaissance detection system (SWORDS) robots that were used by the U.S. military during the Iraq war. It is designed for reconnaissance, surveillance, and target acquisition and can have a 360−degree360−degree view. Depending on the circumstances, MAARS can drape much firepower into its tiny frame. A variety of ammunition such as tear gas, nonlethal lasers, and grenade launcher can be wrapped in it. MAARS is an army robot that can fight autonomously thus reducing risk to soldiers' lives while also protecting itself. This robot has seven types of sensors to track the heat signature of an enemy during the day and night. It uses night vision cameras to monitor enemy activities during the night. On command, MAARS fires at opponents. Its other uses include moving heavy loads from one place to another. It provides a range of options from nonlethal force such as warning of an attack. It can also form a two-sided communication system. The robot can also use less lethal weapons such as laughing gas, pepper spray, and smoke and start clusters to disperse crowds. The robot can be controlled from about one kilometer and is designed to increase or decrease speed, climb stairs, and walk on nonpaved roads using wheels rather than tracks.

Compiled from T. Dupont. (2015, October 15). “The MAARS Military Robot.” Prezi. https://prezi.com/fsrlswo0qklp/the-maars-military-robot/ (accessed September 2018); Modular Advanced Armed Robotic System. (n.d.). Wikipedia. https://en.wikipedia.org/wiki/Modular_Advanced_Armed_Robotic_System (accessed September 2018); “Shipboard Autonomous Firefighting Robot – SAFFiR.” (2015, February 4). YouTube. https://www.youtube.com/watch?time_continue=252&v=K4OtS534oYU (accessed September 2018).

SAFFIR (Shipboard Autonomous Firefighting Robot)

Fire on a ship is one of the greatest risks to shipboard life. Shipboard fires have a different and crucial set of problems. Because of the confined space, there are challenges regarding smoke, gas, and limited ability to escape. Even though procedures like fire drills, onboard alarms, fire extinguishers, and other measures provide ways of dealing with fire on the sea, modern technology is in place to tackle this threat in a better way. A U.S. Navy team at the Office of Naval Research has developed SAFFiR. It is a 5 foot 10 inch tall robot. It is not designed to be completely autonomous. It has a humanoid robotic structure so that it can pass through confined aisles and other nooks and corners of a ship and climb ladders. The robot has been designed to work with the obstacles in the passageways in a ship. SAFFiR can use protective fire gear such as fire-protective coats, suppressants, and sensors that are designed for humans. Lightweight and low-friction linear actuators improve its efficiency and control. It is equipped with several sensors: regular camera, gas, and infrared camera for night vision and in black smoke. Its body is designed not only to be fire resistant but also to throw extinguishing grenades. It can work for around half an hour without needing a charge. SAFFiR can also balance itself on an uneven surface.

Compiled from K. Drummond. (2012, March 8). “Navy’s Newest Robot Is a Mechanized Firefighter.” wired.com. https://www.wired.com/2012/03/firefight-robot/ (accessed September 2018); P. Shadbolt. (2015, February 15). “U.S. Navy Unveils Robotic Firefighter.” CNN. https://www.cnn.com/2015/02/12/tech/mci-saffir-robot/index.html (accessed September 2018); T. White. (2015, February 4). “Making Sailors ‘SAFFiR’ – Navy Unveils Firefighting Robot Prototype at Naval Tech EXPO.” America’s Navy. https://www.navy.mil/submit/display.asp?story_id=85459 (accessed September 2018).

Pepper

Pepper is a semihumanoid robot manufactured by SoftBank Robotics that can understand human emotions. A screen is located on its chest. It can identify frowning, tone of voice, smiling, and user actions such as the angle of a person’s head and crossed fingers. This way Pepper can determine if a person’s mood is good or bad. Pepper can walk autonomously, recognize individuals, and can even lift their mood through its conversation.

Pepper has a height of 120 cms120 cms (about 4 feet4 feet). It has three directional wheels attached, enabling it to move all around the place. It can tilt its head and move its arms and fingers and is equipped with two high-definition cameras to understand the environment. Because of its anticollision functionalities, Pepper reduces unexpected collisions and can recognize humans as well as obstacles nearby. It can also remember human faces and accepts smartphone and card payments. Pepper supports commands in Japanese, English, and Chinese.

Pepper is deployed in service industries as well as homes. It has several advantages for effectively communicating with customers but has also been criticized at places for incompetence or security issues. The following examples provide information on its applications and drawbacks:

· Interacting with robots while shopping is changing the face of AI in commercial settings. Nestlé Japan, a leading coffee manufacturer, has employed Pepper to sell Nescafé machines to enhance customer experience. Pepper can explain the range of products Nestlé has to offer and recognize human responses using facial recognition and sounds. Using a series of questions and responses to them, the robot identifies a consumer’s need and can recommend the appropriate product.

· Some hotels such as Courtyard by Marriott and Mandarin Oriental are employing Pepper to increase customer satisfaction and efficiency. The hotels use Pepper to increase customer engagement, guide guests toward activities that are taking place, and promote their reward programs. Another goal is to collect customer data and fine-tune the communication according to customer preferences. Pepper was deployed steps away from the entry at Disneyland theme park hotels, and it immediately increased customer interactions. Hotels use Pepper to converse with guests while they are checking in or out or to guide them to the spa, gym, and other amenities. It can also inform guests about campaigns and promotions and help staff members avoid the mundane task of enrolling guests in a loyalty program. Customer reactions are largely quite positive in regard to this.

· Central Electric Cooperative (CEC), an electric distribution cooperative located in Stillwater, Oklahoma, has installed Pepper to monitor outages. CEC serves more than 20,000 customers in seven counties in Oklahoma. Pepper is connected to the operations center to read information about live outages, and by connecting them to geographic information system (GIS) maps it can also inform operations about the live locations of service trucks. At CEC, Pepper is also used for conferences where attendees can know more about the company and its services. Pepper answers a range of questions regarding energy consumption. In the future, the company plans to invest more in robots to meet its requirements. See Figure 10.2 that shows Pepper participating as a team member during a prospective employee interview to provide input about CEC’s programs and so on.

· Fabio, a Pepper robot, was installed as a retail assistant at an upmarket food and wine store in England and Scotland. A week after implementing it, the store pulled the service because it was confusing customers, and they preferred the service from personal staff rather than Fabio. It provided generic answers on queries such as the shelf location of items. However, it failed to understand completely what the customer was requesting due to background noise. Fabio was provided another chance by placing it in a specific area that attracted only a few customers. Then they also complained about Fabio’s inability to move around the supermarket and direct them to a specific section. Surprisingly, the staff at the market became accustomed to Fabio rather than considering it as a competitor.

· Pepper has several security concerns that were pointed out by Scandinavian researchers. According to them, it is easy to have unauthenticated root-level access to the bot. They also found the robot to be prone to brute force attack. Pepper’s functions can be programmed using various application programming interfaces (APIs) through languages such as Python, Java, and C + +. This feature can cause it to provide access to all its sensors, making it not secure. An attacker can establish a connection and then use Pepper’s mic, camera, and other features to spy on people and their conversations. This is an ongoing issue for many robots and smart speakers.

Compiled from “Pepper Humanoid robot helps out at hotels in two of the nation’s most-visited destinations (2017)”. SoftBank Robotics. https://usblog.softbankrobotics.com/pepper-heads-to-hospitality-humanoid-robot-helps-out-at-hotels-in-two-of-the-nations-most-visited-destinations (accessed November 2018); R. Chirgwin. (2018, May 29). “Softbank’s ‘Pepper’ Robot Is a Security Joke.” The Register. https://www.theregister.co.uk/2018/05/29/softbank_pepper_robot_multiple_basic_security_flaws/ (accessed September 2018); A. France. (2014, December 1). “Nestlé Employs Fleet of Robots to Sell Coffee Machines in Japan.” The Guardian. https://www.theguardian.com/technology/2014/dec/01/nestle-robots-coffee-machines-japan-george-clooney-pepper-android-softbank (accessed September 2018); Jiji. (2017, November 21). “SoftBank Upgrades Humanoid Robot Pepper.” The Japan Times. https://www.japantimes.co.jp/news/2017/11/21/business/tech/softbank-upgrades-humanoid-robot-pepper/#.W6B3qPZFzIV (accessed September 2018); C. Prasad. (2018, January 22). “Fabio, the Pepper Robot, Fired for ‘Incompetence’ at Edinburgh Store.” IBN Times. https://www.ibtimes.com/fabio-pepper-robot-fired-incompetence-edinburgh-store-2643653 (accessed September 2018).

Da Vinci Surgical System

Over the last decade, the use of robotics has emerged in surgeries. One of the most famous robotic systems used in surgery is the Da Vinci system that has performed thousands of surgeries. According to surgeons, Da Vinci is the most ubiquitous robot used in more units than any other robot. It is designed to perform numerous nominally invasive operations and can perform simple as well as complex and delicate surgeries. The critical components of Da Vinci are the surgeon console, patient side cart, endowrist instruments, and vision system.

The surgeon console is where the surgeon operates the machine. It provides a high-definition, 3D image of the inside of the patient’s body. The console has master controls that a surgeon can grasp by the robotic fingers and operate on the patient. The movements are accurate and in real time, and the surgeon is entirely in control and can prevent the robotic fingers from moving by themselves. The patient side cart is the location where the patient resides during the operation. It has either three or four arms attached that the surgeon controls using master controls, and each arm has certain fixed pivot points around which the arms move. The third component is the endowrist instruments, which are available while performing surgery. They have a total of seven degrees of freedom, and each instrument is designed for a specific purpose. Levers can be released quickly for a change of instruments. The last component is a vision system, which has a high-definition, 3D endoscope and image-processing device that provides real-life images of the patient’s anatomy. A viewing monitor also helps the surgeon by providing a broad perspective during the process.

Patients who have surgery that used the Da Vinci system heal faster than those performed by traditional methods because the cuts by robotic arms are quite small and precise. A surgeon must undergo online and hands-on training and must perform at least five surgeries in front of a surgeon who is certified to use the Da Vinci system. This technology does increase the cost of the surgery, but its ability to ease pain while increasing precision makes it the future of such procedures.

Compiled from “Da Vinci Robotic Prostatectomy – A Modern Surgery Choice!” (2018). Robotic Oncology. https://www.roboticoncology.com/da-vinci-robotic-prostatectomy/ (accessed September 2018); “The da Vinci® Surgical System.” (2015, September). Da Vinci Surgery. http://www.davincisurgery.com/da-vinci-surgery/da-vinci-surgical-system/ (accessed September 2018).

Snoo – A Robotic Crib

Snoo, a robotic, Wi-Fi-enabled crib was developed by Yves Behar, pediatrician Dr. Harvey Karp, and MIT–trained engineers. According to its designers, Snoo mimics Dr. Karp’s famous sleep strategy called the five S’s, which implies swaddled, side or stomach position, shush, swing, and suck. Snoo is an electrified crib that puts babies to sleep automatically. It recreates sensations experienced by the child during the last trimester of a pregnancy. Infants are at maximum ease when they hear white noise, feel movements, and are wrapped, which Snoo provides at par. Once a baby is securely attached to the bassinet, Snoo senses whether it is fussy, keeps track of its movements, and, if found, moves the crib in a womblike motion until the baby calms down. An app can be installed on Snoo’s smartphone to control its speed and white noise. Also, Snoo can be turned off after eight minutes or can continue rocking through the night. The company advertises it as the safest bed ever made with a built-in swaddling strap that ensures that the child does not move from his or her back. Snoo prevents parents from getting up several times in the night to do this themselves; hence, it gives them a sound sleep.

Compiled from S. M. Kelly. (2017, August 10). “A Robotic Crib Rocked My Baby to Sleep for Months.” CNN Tech. https://money.cnn.com/2017/08/10/technology/gadgets/snoo-review/index.html (accessed September 2018); L. Ro. (2016, October 18). “World’s First Smart Crib SNOO Will Help Put Babies to Sleep.” Curbed. https://www.curbed.com/2016/10/18/13322582/snoo-smart-crib-yves-behar-dr-harvey-karp-happiest-baby (accessed September 2018).

MEDi

MEDi, short for Machine and Engineering Designing Intelligence, is available at six hospitals in Canada and one in the United States. MEDi helps reduce stress in children from painful surgeries, tests, and injections. It is two feet tall and weighs around 11 pounds. It looks like a toy. Dr. Tanya Beran proposed using MEDi after working in hospitals where she heard children exclaiming with joy at the sight of the robot. She suggests that since there is not enough pain management expertise available in such situations, technology can provide a helping hand. The robot can speak 19 languages and can easily be integrated into various cultures. Aldebaran built this robot, which calls itself NAO. It can cost $8,000$8,000 and more. Beran bought MEDi to life by adding software that could operate in hospital settings with kids. MEDi strikes up conversations with the kids during a variety of procedures. It was first programmed for flu vaccines and since has been used in other tests. MEDi can even tell story to a child. The robot helps not only children but also nurses by lowering children’s stress and relaxing them. Parents have said that when children leave the hospital, they did not speak about needles and pain but in fact left with happy memories.

Compiled from A. Bereznak. (2015, January 7). “This Robot Can Comfort Children Through Chemotherapy.” Yahoo Finance. https://finance.yahoo.com/news/this-robot-can-comfort-children-through-107365533404.html (accessed September 2018); R. McHugh & J. Rascon. (2015, May 23). “Meet MEDi, the Robot Taking Pain Out of Kids’ Hospital Visits.” NBC News. https://www.nbcnews.com/news/us-news/meet-medi-robot-taking-pain-out-kids-hospital-visits-n363191 (accessed September 2018).

Care-E Robot

Airports are growing in size and the number of people who go to them, and this has increased air traffic, flight cancellations, and gate switches, causing travelers to run to different boarding gates. KLM Royal Dutch Airlines is trying a new way to ease this process from problems related to security, boarding gates, and hectic travel with the use of the blue bot “Care-E Robot.” This service is scheduled to launch at international airports in New York and San Francisco. This robot could be found at security checkpoints and take travelers and their carry-on luggage wherever they need to go. Through its nonverbal sounds and signals, Care-E directs travelers to scan their boarding passes and, once scanned, is at their service when they are busy strolling the shops or using restrooms. Care-E also avoids collision using eight sensors with its “peripheral collision avoidance.” One of its best features is to relate boarding gate changes to travelers and provide them transportation to the newly assigned gate.

Care-E Robot can carry luggage weighing up to 80 pounds. It runs at a speed of 3 mph3 mph, which might be a little too slow for someone running late to catch a flight. However, early travelers who want to explore the airport can use Care-E on a free trial for two days. Implementations of robots like these have not yielded the desired results due to frequent changes in airport policies regarding batteries, but the market for such a robot is quite optimistic about its future.

Compiled from M. Kelly. (2018, July 16). “This Adorable Robot Wants to Make Air Travel Less Stressful.” The Verge. https://www.theverge.com/2018/7/16/17576334/klm-royal-dutch-airlines-robot-travel-airport (accessed September 2018); S. O’Kane. (2018, May 17). “Raden is the Second Startup to Bite the Dust After Airlines Ban Some Smart Luggage.” Circuit Breaker. https://www.theverge.com/circuitbreaker/2018/5/17/17364922/raden-smart-luggage-airline-ban-bluesmart (accessed September 2018).

AGROBOT

The combination of sweetness loaded with multiple health benefits makes strawberries one of the world’s most popular and consumed fruits. Close to 5 million tons of strawberries are harvested every year, an upward trend in the United States, Turkey, and Spain as top harvesters. AGROBOT, a company engaged in the business of agricultural robots, has developed a robot that can harvest strawberries at any place. Robots using 24 robotic manipulators built on a mobile platform work to identify superior quality strawberries.

Strawberries require a high degree of care because they are delicate compared to other fruits. Fruits such as apples, bananas, and mangoes ripen after being picked whereas strawberries are picked at their full maturity. Hence, harvesting strawberries has been an entirely manual process until recently. AGROBOT was developed in Spain; this robot performs automated processes except selecting the strawberries and packing them. To protect strawberries from being squeezed during picking, the robot cuts them with two razor-sharp blades and catches them in baskets lined with rubber rolls. Once full, the baskets are placed on a conveyor belt and passed to the packing station. Human operators can directly select and pack the berries.

AGROBOT is operated by one man, and a maximum of two people can ride on it. Robotic arms control the coordination between blades and basket. The robot has four main components: inductive sensors, ultrasonic sensors, a collision control system, and a camera system. Camera-based sensors view each fruit and analyze it for ripeness according to its form and color; once a berry is ripe, the robot cuts it from its branches with precise movements. Each arm is fortified with two inductive sensors to stop at the end positions. The collision control system must be capable of responding to dust, temperature change, vibration, and shock; hence, an ultrasonic sensor is attached to the robot to prevent the arms from touching the ground. Each wheel is equipped with ultrasonic sensors to determine the distance between the strawberry and the robot’s current position. These sensors also help in keeping the robot on track and preventing damage to the fruit. Signals received from the sensors are continuously transmitted to an automatic steering system to regulate the position of wheels.

Compiled from “Berry Picking at Its Best with Sensor Technology.” Pepperl+Fuchs. https://www.pepperl-fuchs.com/usa/en/27566.htm (accessed September 2018); R. Bogue. (2016). “Robots Poised to Revolutionise Agriculture.” Industrial Robot: An International Journal, 43(5), pp. 45–456; “Robots in Agriculture.” (2015, July 6). Intorobotics. https://www.intorobotics.com/35-robots-in-agriculture/ (accessed September 2018).

Section 10.4 Review Questions

1. Identify applications of robots in agriculture.

2. How could a social support robot such as Pepper or MEDi be useful in healthcare?

3. Based on the illustrative applications of robots in this section, build a matrix where the rows are the robots’ capabilities and the columns are industries. What similarities and differences do you observe across these robots?

10.5 Components of Robots

Depending on their purpose, robots are made of different components. However, all robots have some common ones, and others are tweaked according to a robot’s purpose. Figure 10.3 identifies the components. Common components of the robots are described next.

Figure 10.3 General Components of a Robot.

Figure 10.3 Full Alternative Text

Power Controller

A power controller is the driving force of a robot. Most robots run on batteries, but a few are powered by a direct current (DC) electrical supply. Other factors (i.e., usage, sufficient power to drive all parts) must be kept in mind while designing robots.

Sensors

Sensors are used to direct a robot in its surrounding. Force sensors, ultrasound sensors, distance sensors, laser scanners, and so on help robots to make decisions according to their environment. Sensors are used for robots to identify speech, vision, temperature, position, distance, touch, force, sound, and time. Vision sensors or cameras are used to build a picture of the environment and for the robot to learn about it and to differentiate between which items to choose and which to ignore. In collaborative robots, sensors are also used to prevent them from bumping into humans or other robots. This way, humans and robots can work next to each other without the fear that the robot might unintentionally harm the human. Sensors collect information and send it to the central processing unit (CPU) electronically.

Effectors or Rover or Manipulator

An effector is nothing but a body of a robot. It can also describe the devices that affect the environment, such as hands, legs, arms, bodies, and fingers. The CPU controls the actions of effectors. An essential function of them is to move the robot and other objects from one place to another, and their characteristics depend on the role that has been outlined. Industrial robots have end effectors that contribute to the robot’s work as a hand. Depending on the type of robot, end effectors can be magnets, welding torches, or vacuums.

Navigation or Actuator System

Actuators are devices that define how a robot travels. With the help of an actuator, electrical energy converts into mechanical energy, enabling the robot to move back, forward, left, right and to lift, drop, and perform its job. The actuator can be a hydraulic cylinder or an electric motor. The actuator system is the way that all of the robot’s components are embedded into one.

Controller/CPU

This is the brain of the robot and has the AI embedded in it. The CPU allows a robot to perform its function by connecting all systems into one. It also provides commands for the robot to learn from the surrounding movement of the body or any of its actions.

Section 10.5 Review Questions

1. What are the common components of a robot?

2. What is the function of sensors in a robot?

3. How many different types of sensors might exist in a robot?

4. What is the function of a manipulator?

10.6 Various Categories of Robots

Robots perform a variety of functions. Depending on these, robots can be categorized into the following categories.

Preset Robots

Preset robots are preprogrammed. They have been designed to perform the same task over time and can work 24 hours a day, 7 days a week without any breaks. Preset robots do not alter their behavior. Therefore, these robots have an incredibly low error rate and are suitable for wearisome work. They are frequently used in manufacturing sectors such as the mobile industry, vehicle manufacturing, material handling, and welding to save time and money. Preset robots deliver jobs in environments where it is hazardous for humans to work. Robots move heavy objects, perform assembly tasks, paint, inspect parts, and handle chemicals. A preset robot articulates according to the operation it performs. It can perform a significant role in the medical field because the tasks it performs must have high efficiency at a level comparable to human beings.

Collaborative Robots or Cobots

Cobots are the robots that can collaborate with human workers, assisting them to achieve their goals. The use of cobots is trending in the market, and there is an excellent outlook for collaborative robots. According to the survey by MarketsandMarkets, the cobots market in 2020 will be worth around $3.3$3.3 billion. There are various functions of collaborative robots. Depending on the usage, the collaborative robots are used. Collaborative robots have various applications in manufacturing as well as the medical industry.

Stand-Alone Robots

Stand-alone robots are the robots that have a built-in AI system and work independently without much interference from humans. These robots perform tasks depending on the environment and adapt to changes in it. With the use of AI, a stand-alone robot learns to modify its behavior and excel in performing its assignment. Autonomous robots have household, military, education, and healthcare applications. They can walk like a human being, avoid obstacles, and provide social-emotional support. Some of these robots are used for domestic purposes as stand-alone vacuum cleaners, such as iRobot Roomba. Stand-alone robots are also used in hospitals to deliver medications, keep track of patients who are yet to receive them, and send this information to the nurses working on that shift and other shifts without chance of any error.

Remote-Controlled Robots

Even though robots can perform stand-alone tasks, they do not have human brains; hence, many tasks require human supervision. These robots can be controlled via Wi-Fi, Internet, or satellite. Humans direct remote-controlled robots to perform complicated or dangerous tasks. The military uses these robots to detonate bombs or to act as soldiers around the clock on the battlefield. In the space program research field, their scope of use is extensive. Remote-controlled cobots are also used to perform marginally invasive surgeries.

Supplementary Robots

Supplementary robots enhance the existing capabilities or replace capabilities that a human has lost or does not have. This type of robot can be directly attached to a human’s body. It connects to a user’s body and communicates with the robot’s operator directly or when the operator grips the body. The robot can be controlled by a human body, and in some cases, even by thinking of a specific action. Its applications include serving as a robotic prosthetic arm or providing precision for the surgeons. Extensive research on building prosthetic limbs is being conducted.

Section 10.6 Review Questions

1. Identify some key categories of robots.

2. Define and illustrate the capabilities of a cobot.

3. Distinguish between a preset robot and a stand-alone robot. Give examples of each.

10.7 Autonomous Cars: Robots in Motion

A robot that may eventually touch most people’s lives is an autonomous (self-driving) car. Like many other technologies, self-driving cars have been at peak hype recently, but people also recognize their technical, behavioral, and regulatory challenges. Nevertheless, technology and processes are evolving to make the self-driving car a reality in the future, at least in specific settings if not all over the world. Early versions of self-driving cars were enabled by the radio antenna developed in 1925. In 1989, researchers at Carnegie Mellon used neural networks to control an autonomous vehicle. Since then, many technologies have come together to accelerate development of self-driving cars. These include:

· MOBILE PHONES: With the help of low-powered computer processors and other accessories such as cameras, mobile phones have become ubiquitous. Many technologies developed for phones, such as location awareness and computer vision, are finding applications in cars.

· WIRELESS INTERNET: Connectivity has become much more feasible with the rise of 4G networks and Wi-Fi. Going forward, growth in 5G will perhaps be important for self-driving cars to allow their processors to communicate with each other in real time.

· COMPUTER CENTERS IN CARS: A number of new technologies are available in today’s cars, such as rearview cameras and front and back sensors that help vehicles detect objects in the environment and alert the driver to them or even take necessary actions automatically. For example, adaptive cruise control automatically adjusts the speed of a car based upon the speed of the vehicle in front.

· MAPS: Navigation maps on mobile phones or navigation systems in cars have made a driver’s job easy with regard to navigation. These maps enable an autonomous vehicle to follow a specific path.

· DEEP LEARNING: With advances in deep learning, the ability to recognize an object is a key enabler of self-driving cars. For example, being able to distinguish a person from an object such as a tree, or whether the object is moving or stationary is critical in taking actions in a moving vehicle.

Autonomous Vehicle Development

The heart of an autonomous vehicle system is a laser rangefinder (or light detection and ranging – lidar device), which is on the vehicle’s roof. The lidar generates a 3D image of the car’s surroundings and then combines it with high-resolution world maps to produce different data models for taking action to avoid obstacles and follow traffic rules. In addition, many other cameras are mounted. For example, a camera positioned near a rearview mirror detects traffic lights and takes videos. Before making any navigation decisions, the vehicle filters all data collected from the sensor and camera and builds a map of its surroundings and then precisely locates itself in that map using GPS. This process is called mapping and localization.

The vehicle also consists of other sensors such as the four radar devices that are on the front and back bumpers. These devices allow the vehicle to see far distances so that they can make decisions beforehand and deal with fast-moving traffic. A wheel encoder determines the vehicle’s location and maintains records of its movements. Algorithms such as neural networks, rule-based decision making, and a hybrid approach are used to determine the vehicle’s speed, direction, and position, and the collected data are used to direct the vehicle on the road to avoid obstacles.

Autonomous vehicles must rely on detailed maps of roads. Thus, before sending driverless cars on roads, engineers drive a route several times and collect data about its surroundings. When driverless vehicles are in operation, engineers compare the data acquired by them to the historical data.

There is an entire town built for the sole purpose of testing autonomous vehicles. It is located in Michigan. This city has no single resident, and self-driving vehicles roam the streets without the risks in the real world. This city, called Mcity, is truly a city for robotic vehicles. Mcity includes intersections, traffic signals, buildings, construction work, and moving obstacles such as humans and bicycles similar to those in real cities. Autonomous vehicles are not only tested in this closed environment but are being used in the real world as well.

Google’s Waymo unit is one of the early pioneers of self-driving vehicles. They have been tested on California roads, but before they start to drive next to human-driven cars, companies have to test them thoroughly because one negative incident can impede their acceptance. For example, in the spring of 2018, a self-driving vehicle being tested by Uber killed a pedestrian in Tempe, Arizona. This led to the suspension of all public testing of autonomous vehicles by Uber. The technology is still in development, but it has come far enough that limited testing on public roads is safe. We might be surprised in the near future by the fact that the person in the driver’s seat of a vehicle next to you in traffic might not actually be driving it at all.

In 2016, the U.S. Department of Transportation (DOT) began to embrace driverless vehicles to speed their development. In September 2016, DOT announced the first-ever guidelines for autonomous driving. A groundbreaking announcement by the National Highway Traffic Safety Administration (NHTSA) a month later allowed for the AI system controlling Google’s self-driving vehicle to be considered a driver in response to the company’s proposal to the NHTSA in November 2015.

Some states currently have specific laws that ban autonomous driving. For example, as of this writing, the state of New York does not allow any hands-free driving. Without clear regulations, testing self-driving vehicles is a challenge. Although a few states such as Arizona, California, Nevada, Florida, and Michigan currently allow autonomous vehicles on the road, California is the only one with licensing regulations at this point.

Google might be the most well known for autonomous vehicles, but it is not the only one. A handful of the most powerful companies, such as Uber and Tesla, are in the same race as well. Every major car company is working either with technology companies or its own technology to develop autonomous vehicles or at least to participate in this revolution.

Issues with Self-Driving Cars

Autonomous cars have been connected to a number of issues.

· CHALLENGES WITH TECHNOLOGY: There have been several challenges with the technology used in self-driven cars. Several software and mechanical hurdles are still to be overcome in order to roll out a fully autonomous car. For example, Google is still trying to update its software on an almost daily basis for its self-driven car. Several other companies are still trying to figure out the amount of authority to be transferred when a human driver takes control from an automatic vehicle.

· ENVIRONMENTAL CHALLENGES: Technology and mechanical capabilities cannot yet address many environmental factors affected by self-driving cars. For example, there are still concerns regarding their performance in bad weather. Likewise, several systems have not been tested in extreme conditions such as snow and hail. There are several tricky navigating situations on the road, such as when an animal jumps onto it.

· REGULATORY CHALLENGES: All companies planning to become involved with self-driving cars need to address regulatory hurdles. There are still many unanswered questions about the regulation of autonomous driving. Several questions about liability include these: What will a license involve? Will new drivers be required to get traditional licenses even if they are not drivers? What about young people, or older people with disabilities? What will be required to operate these new vehicles? Governments need to work quickly to catch up with the booming technology. 

· Considering that public safety is on the line, auto regulations should be some of the strictest regulations in the modern world.

· PUBLIC TRUST ISSUES: Most people do not yet believe that an autonomous car can keep them safe. Trust and consumer acceptance are the crucial factors. For example, if there is a situation when an autonomous car is being forced to choose between the life of a passenger versus that of a pedestrian, what should be done? Consumers may refuse the whole idea of driverless cars. No technology can be perfect, but the question is which company will be able to best convince its customers to entrust their lives to them.

Advances similar to those for self-driving cars are being explored in other autonomous vehicles. For example, several companies have already launched trials of self-driving trucks. Autonomous trucks, if ever fully deployed, will have a massive disruptive effect on jobs in the transportation industry. Similarly, self-driving tractors are being tested. Finally, autonomous drones and aircrafts are also being developed. These developments will have a huge impact on future jobs while creating other new jobs in the process.

Self-driving vehicles have become part of this world of technology in spite of related technical and regulatory barriers. Autonomous vehicles are yet to achieve the knowledge capabilities of human drivers, but as the technology improves, more-reliable driving vehicles will become a reality. Like many technologies, the short-term impact may be cloudy, but the long-term impact is yet to be determined.

Section 10.7 Review Questions

1. What are some of the key technology advancements that have enabled the growth of self-driving cars?

2. Give examples of regulatory issues in self-driving cars.

3. Conduct online research to identify the latest developments in autonomous car deployment. Give examples of positive and negative developments.

4. Which type of self-driving vehicles are likely to have the most disruptive effect on jobs, and why?

10.8 Impact of Robots on Current and Future Jobs

Robotics has been a boon to the manufacturing industry. Besides automation that is possible with robotics, new technologies such as image recognition systems are automating jobs that used to require humans for inspection and quality control.

Various industry experts report that by 2025, up to 25 percent of current jobs will be replaced by robots or AI. Davenport and Kirby’s book Humans Need Apply: Winners and Losers in the Age of Smart Machines (2016) focuses on this topic. Of course, many other researchers, journalists, consultants, and futurists have given their own predictions. In this section, we review some related issues. These issues are relevant to AI in general and robotics in particular. Thus, Chapter 14 will also cover these issues, but we want to study these in the context of robotics in this chapter.

As a group activity, watch the following video: https://www.youtube.com/watch?v=GHc63Xgc0-8. Also watch https://www.youtube.com/watch?v=ggN8wCWSIx4 for a different view. What are your takeaways from these videos? What is the most likely scenario in your view? How can you prepare for the day when indeed humans may not need to apply for many jobs?

IBM Watson’s ability to digest vast amounts of data in the medical research literature and provide the latest information to a physician has been written about in the literature. Similar job enhancement opportunities in many other areas have been seen. Consider this: AI-powered technologies such as narrative science and automated insights that can ingest structured data include visualizations generated by software such as Tableau and develop an initial draft of a story to narrate what the results convey. Of course, that would appear to threaten the job of a journalist or even a data scientist. In reality, this can also enhance that job by presenting an initial draft of a story. Then the storyteller can focus on more advanced and strategic issues related to that data and visualization.

The power of consistency and comprehensiveness can also be helpful in the completion of jobs. For example, as noted by Meister (2017), chatbots can likely provide much of the initial human resource (HR) information to new employees. Chatbots can also be helpful in providing such information to remote employees. A chatbot is more likely than a human to provide complete and consistent information each time. Of course, this implies that workers whose main job is to recite such information to each new employee or serve as the first source of information may not be needed.

Hernandez (2018) identified seven job categories into which robotics in particular and AI in general will expand. She also quoted several other studies. According to a McKinsey & Co study, AI could result in 20−50 million20−50 million new jobs in the next 10−15 years.10−15 years. McKinsey also predicts that 75−375 million75−375 million people may need to change jobs/occupations in the same time period because of robotics and AI. According to Hernandez, the following seven jobs are likely to increase:

1. AI DEVELOPMENT: This is an obvious growing area. As more companies develop products and services based on AI, the need for such developers will continue to increase. As an example, iRobot Co, which produces robotic vacuum cleaners, is shifting its hiring from hardware to software engineers as it works to develop its next generation of products that are more adaptive and AI based. Newer robot vacuums are going to be able to “see” a wall. They can also alert the owner to how long the cleanup took and the area swept.

2. CUSTOMER–ROBOT INTERACTIONS: As more companies deploy robots in these organizations, acceptance of such robots by both employees and customers is uncertain. A new job category has emerged to study the interactions between a robot and its coworkers and customers and to retrain the robot or take this information into account in designing the next generation. Clearly, the study of such interactions may enable the use of analytics/data science as well.

3. ROBOT MANAGERS: Although robots might do the bulk of their work in a specific situation, humans will still need to observe them and ensure that the work is progressing as expected. Further, if any unusual conditions arise, a human worker has to be alerted and respond to the situation. This would be true in many settings where the robots are performing the bulk of tasks in areas such as manufacturing. Hernandez (2018) gives an example of Cobalt robots, which work as security guards. These robots alert a human whenever an intruder is detected or they notice anything unusual. Of course, a human robot manager is typically able to supervise many more robots than human workers because the primary role of the manager is to supervise them and respond to unusual situations.

4. DATA LABELERS: Robots or AI algorithms learn from examples. And the more examples they are given, the better their learning can be (see Chapter 5 for a longer description of this issue). For example, image recognition systems in virtually every setting (see Chapter 5 on deep learning for examples) require as many examples as possible to improve those systems’ recognition capability. This is crucial for not just facial recognition but also image applications to detect cancer from X-ray images, weather features from radar images, and so on. It requires that humans view the example images and label them as representing a specific person, feature, or class. This work is tedious and requires humans. Many companies have hired hundreds of human labelers to view the images and tag them appropriately. As such image applications grow, the need for labelers will also increase. These workers are also needed for continuous improvement of the robot or AI algorithms by recording false positives or newer examples.

5. ROBOT PILOTS AND ARTISTS: Robots in general and drones in particular are being used to provide action shots using overhead cameras or angles that would be difficult if not impossible for humans to do. Drones could also be dressed as birds or flowers and provide a unique overview as well as enhance a setting. Similarly, other robots might be dressed in unique outfits to create a cultural ambiance. Such designer/makeup artists are being hired by many companies that provide services for events such as concerts, weddings, and so forth. In addition, drone piloting has become a highly specialized skill for entertainment, commercial, and military applications. These jobs will increase as the applications evolve.

6. TEST DRIVERS AND QUALITY INSPECTORS. Autonomous cars are already becoming reality. With each such automation of vehicles, at least for the foreseeable future, there is a growing need for safety drivers who monitor each vehicle’s performance and take appropriate actions in unusual situations. Their jobs would not entail the use of remote controls as drone pilots employ but continuous watch of the vehicle’s operations and response to emergent situations. Similar jobs also exist in other robotic applications as the robots are trained and tested to work in specific settings.

7. AI LAB SCIENTISTS. This brings us to the very first category of new jobs we identified—AI coders. While their job is to develop the algorithms for robots or AI programs, a similar category of highly specialized users is also emerging—folks who are trained and employed in using these hardware and software systems for special applications. For example, physicians have to undergo additional training to be certified in the use of robots in their surgeries, cardiology and urology practices, and so on. Another category of such specialists involves scientists who customize these robots and AI algorithms for their domain. For example, quite a few companies are using AI tools to identify new drug molecules to develop and test new treatment options for diseases. AI could speed such development. These scientists not only develop their domain expertise but also data scientists’ knowledge or at least the ability to work with data scientists to create their new applications.

Although the preceding list identifies several categories of jobs that are likely to develop or increase, millions of jobs are likely to be eliminated. For example, automation is already impacting the number of jobs in logistics. When autonomous trucks become a reality, at least some of the well-paying jobs in transportation will likely be gone. There might be disagreements on when the massive change may occur, but the long-term impact on jobs is certain to occur. The major issue this time is that many of the knowledge economy “white-collar” jobs are the ones that are more likely to be automated. And this change is unprecedented in history. Many social scientists, economists, and leading thinkers are worried about the upheaval that this next wave of robotic automation will cause, and they are considering various solutions. For example, the concept of universal basic income (UBI) has been proposed. UBI proponents argue that giving every citizen a minimal basic income will ensure that no one goes hungry despite the massive loss of jobs that is likely to occur. Others, for example, Lee (2018), have argued that providing UBI may not satisfy human beings’ need for meaningful achievements and contributions in life. Lee proposes a social investment stipend (SIS), which would recognize individuals’ contributions to society for providing support and care, community service, or education. The stipend would be paid in recognition of an individual’s service in one of these categories. Lee’s book focuses on this issue and is one of the many ideas being proposed on how to plan for and address the upcoming disruption from automation. Our goal in this section is to simply alert you to these issues.

Section 10.8 Review Questions

1. Which jobs are most at risk of disappearing as the result of the new robotics revolution?

2. Identify at least three new categories of jobs that are likely to result in a significant number of new employees.

3. Are the tasks undertaken by data labelers just for one time or longer lasting?

4. Research the concepts of UBI and SIS.

10.9 Legal implications of Robots and Artificial Intelligence

As we noted in the previous section, the impact of AI in general and robotics in particular is far and wide and can be studied both specifically in the context of robots and more broadly for AI. Legal implications of robotics and AI are discussed in this chapter and in Chapter 14. Many legal issues are yet to be untangled as we embrace and employ AI technologies, robots, and self-driving cars. This section highlights some of the key dimensions of legal impacts related to AI. The following material has been contributed by Professor Michael Schuster, assistant professor of Legal Studies at the Spears School of Business at Oklahoma State University. He is a noted expert in legal matters related to AI. He has also published extensively in this area.

Tort Liability

Self-driving cars and other systems controlled by AI represent a Pandora’s box of potential tort liability (where a wrongful act creates on obligation to pay damages to another). Imagine that a motorcyclist is injured when a self-driving car veers into his lane and they collide. This was the alleged event that led to Nilsson v. General Motors (N.D. California, 2018)—a case with the potential to address difficult questions of AI-created tort liability. The suit settled, and thus did not clarify who should pay when someone is injured by an AI-controlled system. Potential candidates for liability include programmers of an AI system, manufacturers of a product incorporating AI, and owners of a product at the time it harmed another. Medical malpractice litigation may similarly be altered by new technologies. As doctors defer some decision making to AI, lawsuits for injurious medical care move from professional liability cases (against the doctor) to product liability (against manufacturers of AI systems). An early example of this phenomenon is the lawsuits over allegedly botched surgeries using the Da Vinci Surgical Systems robot.

Patents

The introduction of AI systems capable of independent or human-assisted invention raises a variety of questions about patenting these creations. Patents have traditionally been granted for novel inventions that would not have been an obvious improvement of a known technology as viewed through the eyes of an average party working in the relevant field. Accordingly, this standard has traditionally asked whether a new technology would have been obvious to a human, but as AI becomes ubiquitous, the scope of what is obvious expands. If the average person in an industry has access to an AI system capable of inventing new things, many improvements on known technologies can become obvious. Since these improvements would then be obvious, they would no longer be patentable. Inventing AI will thus make it harder for a human to get a patent as such technology becomes more commonplace. Moving beyond human inventions, a host of issues arise regarding AI that can independently invent. If a person does not contribute to the invention (but rather merely identifies a goal to be achieved or provides background data), he or she does not satisfy the statutory threshold to be an inventor (and thus, does not qualify for patent ownership). See Schuster (2018). If AI creates an invention without a human inventor, who owns the patent or should a patent be granted at all? Some assert the U.S. Constitution’s mandate that patents can be granted only to “inventors” necessarily requires a human actor, and thus, Congress cannot constitutionally allow the patenting of AI-created technologies. Additional commentators present a variety of policy positions arguing why parties such as the computer’s owner, the AI’s creator, or others should own patents for computer-created inventions. These issues are yet to be resolved.

Property

A basic tenet of U.S. law is a strong protection of property rights. These values extend to corporate entities that can own both real estate and movable property to the exclusion of all others while shareholders retain some allotted portion of the corporation itself. Analysts are presently addressing to what extent property rights should extend to autonomous AI. If a robot were to engage in work for hire, might it be able to make purchases of goods or realty to further its interests? Could an eccentric octogenarian millionaire leave his entire fortune to a loyal robot housekeeper? Topics of this nature will raise a variety of new legal queries, including intestate passage of AI-owned property at its “death” and the standard for death in a nonbiological entity.

Taxation

Robotics and AI will replace a significant number of jobs presently undertaken by humans. Commentators are divided on whether new technologies will create a number of new jobs equal to those replaced by automation. Should the scope of new jobs fall shy of those lost, it is feasible that tax-based issues will be created. A particular concern deals with federal payroll taxes whereby workers and employers pay taxes premised on wages made by the employee. If the aggregate number of workers and net pay are reduced by job automation, the payroll tax base will be reduced. Given that these taxes are important to the sustainability of various government-run safety net programs (e.g., Social Security), payroll tax shortfalls could have significant societal ramifications. In 2017, Bill Gates (Microsoft cofounder) set forth a proposal to tax robots that are used to automate existing human jobs. This new tax would theoretically supplement extant payroll taxes to ensure continued funding for government programs. Commentators are split on the advisability (or need) for such a tax. A common criticism is that taxing robots discourages technological advancement, which is contrary to the accepted policy of encouraging such endeavors. A satisfactory resolution to this debate is yet to be reached.

Practice of Law

Beyond what the law is or should be, AI will have substantial effects within discrete segments of the practice of law. A prime example of this influence is in the area of document review—the part of a lawsuit where litigants evaluate documents provided by their opponents for relevance to the case. Costs associated with this process can be substantial given that some cases entail review of millions of pages by attorneys who bill hundreds of dollars per hour. Corporate clients looking to reduce costs—and law firms seeking a competitive advantage—have adopted (or intend to adopt) AI-based document review systems to minimize the number of billed hours. Similarly, some firms have adopted industry-specific technologies to create competitive advantage. At least one major law firm instituted the use of an AI-driven system to analyze strengths and weaknesses of its clients’ patent portfolios.

Constitutional Law

As the state of AI advances, it continues to move toward “human-level” intelligence. But as it becomes more “personlike,” questions arise regarding whether AI should be afforded rights commonly granted to humans. The U.S. Constitution’s First Amendment provides for freedoms of speech, assembly, and religion, but should rights of this nature extend to AI? For instance, one might argue that these rights preclude the government from dictating what a robot can say (violating its right to free speech). At first blush, this proposition seems far-fetched, but perhaps it is not. On the issue, it is notable that the Supreme Court of the United States recently extended some free speech rights and religious liberties to corporate entities. Accordingly, there is some domestic precedent for affording constitutional protections to nonhuman actors. Further, in 2017, Saudi Arabia granted citizenship to “Sophia,” a humanoid robot created by Hong Kong’s Hanson Robotics in 2015. How this issue will resolve itself (domestically and globally) remains to be seen.

Professional Certification

There are many activities for which humans must receive certification issued by a government or professional organization prior to undertaking that act (e.g., the practice of law or medicine). As the state of AI progresses, AI will increasingly be capable of performing these state-regulated endeavors independent of human engagement. With this in mind, standards must be developed to determine whether an AI technology is capable of providing satisfactory service in regulated professional fields. If an autonomous robot is capable of passing a state’s bar exam, should it be able to give legal advice without human supervision? To the extent that many professional groups require annual training to maintain competence, how will these policies apply to AI technologies? Is there value in requiring that a computer undertaking legal functions “attend” continuing legal education classes? These issues will be settled as AI begins to carry out work currently done exclusively by human professionals like doctors and lawyers.

Law Enforcement

In addition to policy choices detailing what the law is or should be, AI may influence enforcement of the law. Rapid growth in technology will soon afford police forces access to large amounts of near real-time data and computing capacity to determine where crimes are being committed. Recognized infractions may run the gamut from common public transgressions (e.g., running a red light) to more private acts, such as underreporting income on a tax return. The capacity to recognize such criminal acts on a large scale raises a variety of enforcement questions. Discretion in prosecuting infractions has long been a part of law enforcement. Should this power of choice regarding issues associated with stereotype-based prosecution decisions be delegated to AI systems? Moreover, machine-based enforcement programs have consistently been met with questions of their constitutionality (e.g., using cameras to identify drivers not stopping for red lights). While these arguments have thus far proven unsuccessful, they will likely be relitigated as the practice expands. Beyond enforcement questions, some have raised the possibility of implementing AI in the judiciary. For instance, it has been proposed that data-based sentencing may more successfully achieve targeted goals (e.g., successful education while incarcerated or avoidance of recidivism) than arguably idiosyncratic members of the judiciary. Such a mechanism will, of course, raise potential transparency issues and arguments relating to granting too much power to AI systems.

Regardless of the issues that the last two sections have raised, robotics technologies and applications are evolving rapidly. As managers, you have to continue to think about how to manage these technologies while being fully aware of immediate behavioral and legal issues in implementing the technologies.

Section 10.9 Review Questions

1. Identify some of the key legal issues for robotics and AI.

2. Liability for harm (tort liability) is an obvious early question for any technology. What are some of the key challenges in identifying such liability?

3. Recent news about illegal intervention in elections has led to the discussion about who is responsible for damage control. When chatbots and automated social media systems have the ability to propagate “fake news,” who should be required to monitor them and prevent such action?

4. What are some of the law enforcement issues in employing AI?

Chapter Highlights

· Industrial automation brought the first wave of robots, but now the robots are becoming autonomous and finding applications in many areas.

· Robotic applications span industries such as agriculture, healthcare, and customer service.

· Social robots are emerging as well to provide care and emotional support to children, patients, and older adults.

· All robots include some common components: power unit, sensors, manipulator/effector, logic unit/CPU, and location sensor/GPS.

· Collaborative robots are evolving quickly, leading to a category called cobots.

· Autonomous cars are probably the first category of robots to touch most consumers.

· Self-driving cars are challenging the limits of AI innovation and legal doctrines

· Millions of jobs are at risk of being lost due to the use of robots and AI, but some new job categories will emerge.

· Robots and AI are also creating new challenges in many legal dimensions.

Questions for Discussion

1. Based upon the current state of the art of robotics applications, which industries are most likely to embrace robotics? Why?

2. Watch the following two videos: https://www.youtube.com/watch?v=GHc63Xgc0-8 and https://www.youtube.com/watch?v=ggN8wCWSIx4 for a different view on impact of AI on future jobs. What are your takeaways from these videos? What is the more likely scenario in your view? How can you prepare for the day when humans indeed may not need to apply for many jobs?

3. There have been many books and opinion pieces written about the impact of AI on jobs and ideas for societal responses to address the issues. Two ideas were mentioned in the chapter – UBI and SIS. What are the pros and cons of these ideas? How would these be implemented?

4. There has been much focus on job protection through tariffs and trade negotiation recently. Discuss how and why this focus may or may not address the job changes coming due to robotics and AI technologies.

5. Laws rely on incentive structures to encourage prosocial behavior. For example, criminal law encourages compliance by punishing those who break the law. Patent law incentivizes creation of new technologies by offering inventors a period of limited monopoly during which they can exclusively use their invention. To what extent do these (and other) incentives make sense when applied to AI? How can incentive structures be created to encourage AI devices to behave in prosocial manners?

6. To what extent do extralegal considerations come into play with regard to the above issues? Are there moral (or religious) dimensions to be considered when determining whether AI should be given rights similar to those of a person? Would AI-assisted law enforcement or court action erode faith in the criminal justice system and judiciary?

7. Adopting policies that maximize the value of AI encourages future development of these technologies. Such a course, however, is not without drawbacks. For instance, determining that a “robot tax” is not a preferred policy choice would increase the incentive to adopt a robot workforce and improve any relevant technologies. Elevating the state of robotics is a laudable goal, but in this instance, it would come at the anticipated cost of reduced public funds. How should trade-offs such as these be evaluated? Where should encouragement of technological progress (especially regarding AI) fall in the hierarchy of government priorities?

Exercises

1. Identify applications other than those discussed in this chapter where Pepper is being used for commercial and personal purposes.

2. Go through specifications of MAARS at https://www.qinetiq-na.com/wp-content/uploads/brochure_maars.pdf. What are the functions of MAARS?

3. Conduct online research to find at least one new robotics application in agriculture. Prepare a brief summary of your research: the problem addressed, technology summary, results achieved if any, and lessons learned.

4. Conduct online research to find at least one new robotics application in healthcare. Prepare a brief summary of your research: the problem addressed, technology summary, results achieved if any, and lessons learned.

5. Conduct online research to find at least one new robotics application in customer service. Prepare a brief summary of your research: the problem addressed, technology summary, results achieved if any, and lessons learned.

6. Conduct online research to find at least one new robotics application in an industry of your choice. Prepare a brief summary of your research: the problem addressed, technology summary, results achieved if any, and lessons learned.

7. Conduct research to identify the most recent developments in self-driving cars.

8. Conduct research to learn and summarize any new investments and partnerships in self-driving cars.

9. Conduct research to identify any recent examples of legal issues regarding self-driving cars.

10. Conduct research to identify any other new types of jobs that would be enabled by AI and robotics beyond what was covered in the chapter.

11. Conduct research to report on the latest projections for job losses due to robotics and AI.

12. Identify case stories for each of the legal dimensions identified by Schuster (2018) in Section 10.9.

References

· “A Brief History of Robotics since 1950.” Encyclopedia.com. http://www.encyclopedia.com/science/encyclopedias-almanacs-transcripts-and-maps/brief-history-robotics-1950 (accessed September 2018).

· Ackerman, E. (2016). IEEE Spectrum. http://spectrum.ieee.org/automaton/robotics/home-robots/tega-mit-latest-friendly-squishable-social-robot (March 5, 2017).

· “Adidas’s High-Tech Factory Brings Production Back to Germany.” (2017, January 14). The Economist. https://www.economist.com/business/2017/01/14/adidass-high-tech-factory-brings-production-back-to-germany (accessed September 2018).

· Allinson, M. (2017, March 4). “BMW Shows Off Its Smart Factory Technologies at Its Plants Worldwide.” Robotics and Automation. https://roboticsandautomationnews.com/2017/03/04/bmw-shows-off-its-smart-factory-technologies-at-its-plants-worldwide/11696/ (accessed September 2018).

· Aoki, S., et al. (1999). “Automatic Construction Method of Tree-Structural Image Conversion Method ACTIT.” Journal of the Institute of Image Information and Television Engine, 53(6), pp. 888–894 (in Japanese).

· “A Robot Cooks Burgers at Startup Restaurant Creator.” (2018). Techcrunch. https://techcrunch.com/video/a-robot-cooks-burgers-at-startup-restaurant-creator/ (accessed September 2018).

· Ayres, R., & S. Miller. (1981, November). “The Impacts of Industrial Robots.” Report CMU-RI-TR-81-7. Pittsburgh, PA: The Robotics Institute at Carnegie Mellon University.

· Bereznak, A. (2015, January 7). “This Robot Can Comfort Children Through Chemotherapy.” Yahoo Finance. https://finance.yahoo.com/news/this-robot-can-comfort-children-through-107365533404.html (accessed September 2018).

· “Berry Picking at Its Best with Sensor Technology.” (2018). Pepperl+Fuchs. https://www.pepperl-fuchs.com/usa/en/27566.htm (accessed September 2018).

· Bogue, R. (2016). “Robots Poised to Revolutionise Agriculture.” Industrial Robot: An International Journal, 43(5), pp. 450–456

· Broekens, J., M. Heerink, & H. Rosendal. (2009). “Assistive Social Robots in Elderly Care: A Review.” Gerontechnology, 8, pp. 94–103 doi: 10.4017/gt.2009.08.02.002.00.

· Carlsson, B. (1998) “The Evolution of Manufacturing Technology and Its impact on Industrial Structure: An International Study.” IUI Working Paper 203. Internation Joseph A. Schumpeter Society Conference on Evolution of Technology and Market in an International Context. The Research Institute of Industrial Economics (IUI), Stockholm, May 24–28, 1988.

· “Case Study Pepper, Courtyard Marriott.” SoftBank Robotics. https://www.softbankrobotics.com/us/solutions/pepper-marriott (accessed September 2018).

· Chirgwin, R. (2018, May 29). “Softbank’s ‘Pepper’ Robot Is a Security Joke.” The Register. https://www.theregister.co.uk/2018/05/29/softbank_pepper_robot_multiple_basic_security_flaws/ (accessed September 2018).

· Coxworth, B. (2018, May 29). “Restaurant Keeps Its Prices Down – With a Robotic Kitchen.” New Atlas. https://newatlas.com/spyce-restaurant-robotic-kitchen/54818/ (accessed September 2018).

· “Da Vinci Robotic Prostatectomy – A Modern Surgery Choice!” Robotic Oncology. https://www.roboticoncology.com/da-vinci-robotic-prostatectomy/ (accessed September 2018).

· Drummond, K. (2012, March 8). “Navy’s Newest Robot Is a Mechanized Firefighter.” wired.com. https://www.wired.com/2012/03/firefight-robot/ (accessed September 2018).

· Dupont, T. (2015, October 15). “The MAARS Military Robot.” Prezi. https://prezi.com/fsrlswo0qklp/the-maars-military-robot/ (accessed September 2018).

· Engel, J. (2018, May 3). “Spyce, MIT-Born Robotic Kitchen Startup, Launches Restaurant: Video.” Xconomy. https://www.xconomy.com/boston/2018/05/03/spyce-mit-born-robotic-kitchen-startup-launches-restaurant-video/ (accessed September 2018).

· Fallon, S. (2015). “A Blue Robotic Bear to Make Sick Kids Feel Less Blue.” https://www.wired.com/2015/03/blue-robotic-bear-make-sick-kids-feel-less-blue/ (accessed August 2018).

· Forrest, C. (2015).“Chinese Factory Replaces 90%90% of Humans with Robots, Production Soars.” TechRepublic. https://www.techrepublic.com/article/chinese-factory-replaces-90-of-humans-with-robots-production-soars/ (accessed September 2018).

· France, A. (2014, December 1). “Nestlé Employs Fleet of Robots to Sell Coffee Machines in Japan.” The Guardian. https://www.theguardian.com/technology/2014/dec/01/nestle-robots-coffee-machines-japan-george-clooney-pepper-android-softbank (accessed September 2018).

· Gandhi, A. (2013, February 23). “Basics of Robotics.” Slideshare. https://www.slideshare.net/AmeyaGandhi/basics-of-robotics (accessed September 2018).

· Goris, K., et al. (2010, September). “Mechanical Design of the Huggable Robot Probo.” Robotics & Multibody Mechanics Research Group. Brussels, Belgium: Vrije Universiteit Brussels.

· Green, D. (2018). “Adidas Just Opened a Futuristic New Factory – and It Will Dramatically Change How Shoes Are Sold.” Business Insider. http://www.businessinsider.com/adidas-high-tech-speedfactory-begins-production-2018-4 (accessed September 2018).

· Hernandez, D. (2018). “Seven Jobs Robots Will Create – or Expand.” The Wall Street Journal. https://www.wsj.com/articles/seven-jobs-robots-will-createor-expand-1525054021 (accessed September 2018).

· History of Robots. (n.d.). Wikipedia. https://en.wikipedia.org/wiki/History_of_robots (accessed September 2018).

· “Huggable Robot Befriends Girl in Hospital.” YouTube video. https://youtu.be/UaRCCA2rRR0 (accessed August 2018).

· “Innovative Human-Robot Cooperation in BMW Group Production.” (2013, October 9). BMW Press Release. https://www.press.bmwgroup.com/global/article/detail/T0209722EN/innovative-human-robot-cooperation-in-bmw-group-production?language=en (accessed September 2018).

· Javelosa, J., & K. Houser. (2017). “Production Soars for Chinese Factory Who Replaced 90%90% of Employees with Robots.” Future Society. https://futurism.com/2-production-soars-for-chinese-factory-who-replaced-90-of-employees-with-robots/ (accessed September 2018).

· Jeong, S., et al. (2015). “A Social Robot to Mitigate Stress, Anxiety, and Pain in Hospital Pediatric Care.” Proceedings of the Tenth Annual ACM/IEEE International Conference on Human-Robot Interaction Extended Abstracts.

· Jeong, S., et al. (2015). “Designing a Socially Assistive Robot for Pediatric Care.” Proceedings of the Fourteenth International Conference on Interaction Design and Children. ACM.

· Jeong, S., & D. Logan. (2018, April 21–26).“Huggable: The Impact of Embodiment on Promoting Socio-emotional Interactions for Young Pediatric Surgeons.” MIT Media Lab, Cambridge, MA, CHI 2018, Montréal, QC, Canada.

· Jiji. (2017, November 21). “SoftBank Upgrades Humanoid Robot Pepper.” The Japan Times. https://www.japantimes.co.jp/news/2017/11/21/business/tech/softbank-upgrades-humanoid-robot-pepper/#.W6B3qPZFzIV (accessed September 2018).

· Joshua, J. (2013, February 24). “The 3 Types of Robots.” Prezi. https://prezi.com/iifjw387ebum/the-3-types-of-robots/ (accessed September 2018).

· Kelly, M. (2018, July 16). “This Adorable Robot Wants to Make Air Travel Less Stressful.” The Verge. https://www.theverge.com/2018/7/16/17576334/klm-royal-dutch-airlines-robot-travel-airport (accessed September 2018).

· Kelly, S. M. (2017, August 10). “A Robotic Crib Rocked My Baby to Sleep for Months.” CNN Tech. https://money.cnn.com/2017/08/10/technology/gadgets/snoo-review/index.html (accessed September 2018).

· Lee, K. F. (2018). “The Human Promise of the AI Revolution.” The Wall Street Journal. https://www.wsj.com/articles/the-human-promise-of-the-ai-revolution-1536935115 (accessed September 2018).

· Mayank. (2012, June 18). “Basic Parts of a Robot.” maxEmbedded.com. http://maxembedded.com/2012/06/basic-parts-of-a-robot/ (accessed September 2018).

· McHugh, R., & J. Rascon. (2015, May 23). “Meet MEDi, the Robot Taking Pain Out of Kids’ Hospital Visits.” NBC News. https://www.nbcnews.com/news/us-news/meet-medi-robot-taking-pain-out-kids-hospital-visits-n363191 (accessed September 2018).

· Meister, J. (2017), “The Future Of Work: How Artificial Intelligence Will Transform The Employee Experience,” https://www.forbes.com/sites/jeannemeister/2017/11/09/the-future-of-work-how-artificial-intelligence-will-transform-the-employee-experience/ (accessed November 2018).

· Modular Advanced Armed Robotic System. (n.d.). Wikipedia. https://en.wikipedia.org/wiki/Modular_Advanced_Armed_Robotic_System (accessed September 2018).

· Nagato, T., H. Shibuya, H. Okamoto, & T. Koezuka. (2017, July). “Machine Learning Technology Applied to Production Lines: Image Recognition System.” Fujitsu Scientific & Technical Journal, 53(4).

· O’Kane, S. (2018, May 17). “Raden Is the Second Startup to Bite the Dust After Airlines Ban Some Smart Luggage.” Circuit Breaker. https://www.theverge.com/circuitbreaker/2018/5/17/17364922/raden-smart-luggage-airline-ban-bluesmart (accessed September 2018).

· Park, H. W., et al. (2017). “Growing Growth Mindset with a Social Robot Peer.” Proceedings of the Twelfth ACM/IEEE International Conference on Human Robot Interaction.

· Personal Robots Group. (2016). https://www.youtube.com/watch?v=sF0tRCqvyT0 (accessed March 5, 2017).

· Personal Robots Group, MIT Media Lab. (2017). “Growing Growth Mindset with a Social Robot Peer.” Proceedings of the Twelfth ACM/IEEE International Conference on Human Robot Interaction.

· Prasad, C. (2018, January 22). “Fabio, the Pepper Robot, Fired for ‘Incompetence’ at Edinburgh Store.” IBN Times. https://www.ibtimes.com/fabio-pepper-robot-fired-incompetence-edinburgh-store-2643653 (accessed September 2018).

· Ro, L. (2016, October 18). “World’s First Smart Crib SNOO Will Help Put Babies to Sleep.” Curbed. https://www.curbed.com/2016/10/18/13322582/snoo-smart-crib-yves-behar-dr-harvey-karp-happiest-baby (accessed September 2018).

· “Robotics Facts.” Idaho Public Television. http://idahoptv.org/sciencetrek/topics/robots/facts.cfm (accessed September 2018).

· “Robots in Agriculture.” (2015, July 6). Intorobotics. https://www.intorobotics.com/35-robots-in-agriculture/ (accessed September 2018).

· “Robotics: Types of Robots.” ElectronicsTeacher.com. http://www.electronicsteacher.com/robotics/type-of-robots.php (accessed September 2018).

· Rosencrance, L. (2018 May 31). “Tabletop Grapes to Get Picked by Robots in India, with Help from Virginia Tech.” Robotics Business Review. https://www.roboticsbusinessreview.com/agriculture/tabletop-grapes-picked-robots-india-virginia-tech/ (accessed September 2018).

· Schuster, W. M. (2018). “Artificial Intelligence and Patent Ownership.” Washington & Lee L. Rev., 75.

· Shadbolt, P. (2015, February 15). “U.S. Navy Unveils Robotic Firefighter.” CNN. https://www.cnn.com/2015/02/12/tech/mci-saffir-robot/index.html (accessed September 2018).

· “Shipboard Autonomous Firefighting Robot – SAFFiR.” (2015, February 4). YouTube. https://www.youtube.com/watch?time_continue=252&v=K4OtS534oYU (accessed September 2018).

· Simon, M. (2018, May 17). “The Wired Guide to Robots.” Wired. https://www.wired.com/story/wired-guide-to-robots/ (accessed September 2018).

· “Tabletop Grapes to Get Picked by Robots in India.” Agtechnews.com. http://agtechnews.com/Ag-Robotics-Technology/Tabletop-Grapes-to-Get-Picked-by-Robots-in-India.html (accessed September 2018).

· “The da Vinci® Surgical System.” (2015, September). Da Vinci Surgery. http://www.davincisurgery.com/da-vinci-surgery/da-vinci-surgical-system/ (accessed September 2018).

· “Types of Robots.” (2018). RoverRanch. https://prime.jsc.nasa.gov/ROV/types.html (accessed September 2018).

· Westlund, J. K., J. M. Lee, J. Plummer, L. Faridia, F. Gray, J. Berlin, M. Quintus-Bosz, H. Harmann, R. Hess, M. Dyer, S. dos Santos, K. Adalgeirsson, S. Gordon, G. Spaulding, S. Martinez, M. Das, M. Archie, M. Jeong, & C. Breazeal, C. (2016). “Tega: A Social Robot.” Video Presentation. Proceedings of the Eleventh ACM/IEEE International Conference on Human Robot Interaction.

· White, T. (2015, February 4). “Making Sailors ‘SAFFiR’ – Navy Unveils Firefighting Robot Prototype at Naval Tech EXPO.” America’s Navy. https://www.navy.mil/submit/display.asp?story_id=85459 (accessed September 2018).

· Zimberoff, L. (2018, June 21). “A Burger Joint Where Robots Make Your Food.” https://www.wsj.com/articles/a-burger-joint-where-robots-make-your-food-1529599213 (accessed September 2018).

Book: Turban, R.S.D.D. E. (2019). Analytics, Data Science, & Artificial Intelligence. [Yuzu]. Retrieved from https://reader.yuzu.com/#/books/9780135172940/