Workpackage 3 – I-Cave

For successfully realizing cooperative driving solutions for dedicated areas, such as business parks, leisure sites, or event sites, efficient management of fleets of cooperative vehicles is crucial. This project considers the challenges at logistics and system level involved for organizations owning fleets of cooperative vehicles. Organizations that own pools of vehicles for carrying out transport tasks need methods that can support them in determining how to use such pools most efficiently. This relates to dimensions that embrace the whole spectrum of logistics systems: fleet size selection; vehicle rostering and assignment; dispatching; repositioning; and maintenance. Without adequate fleet management methods, it will be extremely hard for such organizations to determine how many vehicles to use when and where, resulting in suboptimal quality of service, possibly too high operational costs, and unawareness about environmental impact / energy consumption. This project focuses explicitly on what the impact and potential is of various fleet management strategies on operational performance of logistical service providers.

The overall goal of this project is to develop efficient methods for the robust management of fleets of cooperative (automated) vehicles by owners of such fleets. Fleet management for cooperative (automated) vehicles is fundamentally different from fleet management for conventional vehicles. First of all, vehicles inside a fleet are information responsive, and can actively exchange information with each other. This includes not only running vehicles (V2V) but also vehicles in standby or parked modes. In addition, the vehicles in a fleet all have computing power available for carrying out calculations (for themselves, or possibly for others); they can make estimates of service plans, and provide updates on actual traffic situations. In this project, we propose a cooperative fleet management structure, in which all fleet components (viz., (dual-cycle) vehicles, cooperative vehicle controllers, and automated fleet managers) work together to obtain the best performance. They do this in environments where other vehicles that are not part of fleets (i.e., regular, non-responsive vehicles) can also be present. At the lowest level, vehicles carry out the actual transport tasks. Groups of vehicles are hereby closely controlled by cooperative vehicle controllers. Information flows in the proposed architecture among all components, and all levels: between vehicles, between vehicles and cooperative vehicle controllers, and between vehicles, vehicle controllers, and fleet managers. The question is then what methods and information exchange protocols should be adopted to obtain the best performance, in this project from a logistical fleet management perspective.

Scientific questions:

What is the best way to estimate and model the (expected) transport demands?
What are the key performance indicators that determine when fleets operate at a high-quality level (from a human / non-human cargo / fleet owner’s perspective)?
In which way are the dynamics of cooperative (automated) fleets, cooperation strategies, and interaction with non-cooperative vehicles best captured in tractable models?
How can the transport demand models, fleet performance indicators, and fleet dynamics models be used to determine efficiently satisfactory fleet operations in real-time?

Staff:

R. R. Negenborn

Prof. dr.