CUSTOMIZE: Customer-driven prescriptive analytics for logistics planning

Samenvatting

B: Intended consortium composition:
Industrial partners: Takeaway, Post NL, Delft University of Technology , Erasmus University of Rotterdam.

C: Short project description.
The proposed research introduces a novel decision-making framework to address the challenges faced by on-demand logistics systems. Having a precise demand forecast is the key success for the operation of any logistics systems. So far, the demand prediction for these systems is estimated at the aggregated level using conventional forecasting models (e.g. time-series, regression). Thus, these approaches cannot provide an insight into the demand at micro-level (e.g. customer behaviour). In other industries such as airlines, it has been shown that forecasting demand at micro-level can greatly improve their operations and revenue. However, this has not been explored in the logistics domain mostly because of data absence.
In this proposal, we first start by developing a demand model to forecast logistics demand at the micro-level. We look to infer the behaviour of customers by using the transaction data which is mostly available by logistics companies. We aim to enhance the predictability of our model by using open data which is publicly available. Thereafter, we integrate our demand forecasting tool into the prescriptive model. The prescriptive model determines the expected number of resources required to perform the operation. In addition, it prescribes various strategies to position resources in a real-time.

D: Connection to the new action agenda of top sector Logistics (Data-Driven logistics)
This research addresses the research agenda of data-driven logistics. The forecasting model is founded by analysing and process the transaction data to infer the customers' behaviour. The enhancement of forecasting prediction by using the open data provides a unique and low-cost solution for the companies. The development of the prescriptive model is an innovative solution to bridge data-analytics and systems operation.
In this project, we will apply our method in two used cases provided by the consortium partners. Our used cases address the research theme of agro logistics and innovation support in the logistics process in operational services. Takeaway is an on-demand logistics provider which is operated in the food industry. Our developed approach uses real-time data and provides a planning approach for its operation. Our second used case focuses on providing home healthcare supply which is provided by health-care organizations. Post-NL recently establishes a logistic services to provide medical devices from pharmacies and medical centres to end-user. Currently, they focus on the patient-centric logistic approach to improve the satisfaction of end-users. The proposed forecasting model and the resolution approach enable them to improve their services taking into account the behavioural aspect of the patients (e.g. satisfaction from the logistics service) and promote the usage of home healthcare services. Their used case addresses the research theme of innovative support in logistics process in operational services as it provides a novel logistics solution for the health care organizations to improve their logistics services.

Kenmerken

Projectnummer

439.19.618

Hoofdaanvrager

Dr. M.Y. Maknoon

Verbonden aan

Technische Universiteit Delft, Faculteit Techniek, Bestuur en Management

Looptijd

01/04/2020 tot 31/12/2021