SINERGI is a four-year project conducted by a unique consortium led by TU Delft, CBS, Tongji University, and Tsinghua University. It will be executed as an integrated program consisting of five work packages: WP1 to WP5.
1: Designing mini-hubs/parking maps
WP1 provides the theoretical basis for the network design of a micro-delivery system, which serves as the foundation for the rest of the project activities. In collaboration with our partners, this work package aims to systematically explore feasible locations for mini-hubs/parking spots in urban, peri-urban, and on-shore areas in four cities. One part of WP1 focuses on the location planning of fixed mini-hubs/parking spots on land, while the other part investigates the potential use of autonomous waterborne vessels as mobile mini-hubs or for redistributing empty (e)bikes through waterways, reducing reliance on heavy vehicles on roads.
2: Modeling rider and consumer behavior
This package is divided into two parts. The first part focuses on understanding rider behavior and identifying the factors that contribute to their riding patterns. The findings from WP2 will be incorporated into WP4 for real-time matching and dispatching. The second part aims to investigate consumer preferences by analyzing their sensitivity to price and delivery time. Surveys will be conducted for micro-delivery companies in the Netherlands and China. The insights gained from this research will inform WP3 in determining the optimal product assortment for each request. Surveys will be designed for two purposes: 1) understanding rider behavior, and 2) capturing consumer preferences.
3: Integrated online product assortment and display optimization
The order in which services are displayed to a new request can influence customer purchase behavior. This WP addresses two aspects: grocery delivery and food delivery. For the first part, we introduce a novel model called the Product Assortment and Display Optimization (PADO), which applies to both offline and online scenarios. The food delivery part of the WP incorporates customer preferences (delivery time and price) based on the findings from WP2. Rider-request matching is performed while respecting customer preferences.
4: Real-time matching and dispatching
WP4 focuses on real-time matching and dispatching decisions by collecting and incorporating real-time traffic data from bike lanes in Copenhagen and Amsterdam. The impact of this integration is evaluated using simulated data based on real traffic information, aiming to determine if this model can reduce the expected number of accidents. The insights from WP2 regarding rider behavior are taken into consideration for these real-time decisions.
5: Knowledge utilization for use cases
This work package consolidates and evaluates the implementation of the innovative models developed in the previous WPs for use cases in Amsterdam, Copenhagen, Singapore, and Shanghai. The results will be disseminated through WP6, involving all main academic collaborators and cooperation partners. Guidelines, recommendation policies, and a comprehensive comparative study will be conducted to share the lessons learned among all partners.