research paper-2

Computational models

Name of the Student: Gopi Krishna Chandragiri

Name of the University: Harrisburg University

Student Id : 222017

Using customer-related data to enhance e-grocery home delivery

The establishment of E-grocery home delivery has brought about a lot of benefits whereby customers receive grocery products at their doorsteps after ordering for them online. Sellers have also managed to make a lot of profits as a result of this innovative technology. However, despite the efforts of sellers to offer reliable and timely delivery services to their customers many deliveries were becoming ineffective due to customers’ absence, and this was causing them to incur substantial operational losses specially on perishable products (Pan, & Qiao, 2017). This study purposes to recommend a computational model that can be used to estimate the probability of customers’ absence in order to optimize the delivery process.

VRP Model

This computational model was built in two stages. The first stage aimed at collecting customer’s related data in order to assess their frequency of purchasing E-grocery products. The researchers established a data mining model whereby they mined data relating to how homes consume facilities such as electricity and water in their homes. By mining this data, the researchers were hoping to understand the frequency of home attendance and this information would help them in assessing the probability of homesteads purchasing groceries online. Also, they worked under the assumption that the frequency of home attendance relates directly to how often they purchase groceries online. In order to arrive at quality results, the researchers worked in collaboration with experts in the various fields such as electrical engineers and water engineers. After collecting all the relevant data, the researchers created statistical correlations between electric power usages by homesteads with their social-economic conditions. This stage aimed at evaluating how consumer’s everyday behaviors relate to energy consumption which would also relates to how often they do online shopping (Fernández-Delgado, & Amorim, 2014). This strategy would help the sellers to know their target customers, and thus save on transportation expenses which used to be incurred in case a customer was absent.

The second stage involved optimizing transportation basing on the data collected. E-grocery home delivery organizations were experiencing a lot of operational expenses when they find that the customer they were delivering products is not available. They used VRP model to relate the number of unsuccessful deliveries with the number of online orders (Toth, & Vigo, D. (Eds.). 2014). The statistical data which they arrived at showed that unsuccessful deliveries were frequent on weekends which implied that people spent less time at homes during the weekends. E-grocery organizations can thus use this computational model to estimate the probability of deliveries being unsuccessful, and consequently they will be able to schedule their transportation plans more effectively (Giannikas, E. 2017). For instance, considering the data they used, they found that most deliveries were unsuccessful on weekends and thus the organizations can reduce their expenditure on transporting goods on weekends.

Though the proposed approach was successful, it has several limitations. The approach relies on the quality of available information and this may not be always accessible due to legal concerns. Data protection from third parties may cause the approach unsuccessful since the necessary data may not be disclosed (Giannikas, E. 2017). Also, the e-grocery home delivery firms provide a time slot which requires customers to specify when they want the order delivered. The approach may thus be unsuccessful since they will lose the specific customer if they fail to deliver the products as agreed. Besides, this data-driven strategy may be ineffective due to computational errors. Just like any other mathematical problem, errors occur and thus if corrections are not made in time the approach may be unsuccessful.

References

Fernández-Delgado, M., Cernadas, E., Barro, S., & Amorim, D. (2014). Do we need hundreds of classifiers to solve real world classification problems? The Journal of Machine Learning Research, 15(1), 3133-3181.

Giannikas, E. (2017). Using customer-related data to enhance e-grocery home delivery.

Pan, S., Giannikas, V., Han, Y., Grover-Silva, E., & Qiao, B. (2017). Using customer-related data to enhance e-grocery home delivery. Industrial Management & Data Systems, 117(9), 1917-1933.

Toth, P., & Vigo, D. (Eds.). (2014). Vehicle routing: problems, methods, and applications. Society for Industrial and Applied Mathematics.