Assessing the Role of AVL in Demand Responsive Transportation Systems

Many-to-many demand responsive transportation systems consist of vehicles which take passengers from their origins to their destinations within a service area. In dial-a-vehicle systems, in order to circumvent the undesirable feature of taxicab systems, vehicles are allowed to deviate from their direct route to serve other passengers and the emphasis is on building efficient tours to increase vehicle productivity. This strategy increases riding times but also increases average occupancy and productivity of the vehicles, and hence decreases average waiting times. A similar problem is faced by the recently developed 'webvan' food delivery service which takes orders for groceries over the internet and commits to delivery to the order's home (or specified address) within a given time frame. The difference is there is a single origin for a defined market area but multiple destinations. It is possible to organize dial-a-transit system in a similar manner but it is not clear how this form of delivery configuration would impact the productive use of the vehicles, drivers and other factors used. The interest we have in this area is to develop a model that can be used to assess the improvement in productive efficiency or of consumer welfare with the use of ITS applications, such as AVL in providing this service. AVL provides information to both the service provider and the service consumer. The information on the supplier side allows the dispatcher to allocate vehicles to achieve some objective. In most cases this has been defined to minimize costs or maximize productivity. We find this to be a narrow definition, after all this is a service industry and meeting the needs of customers should be the objective. This can be defined as maximizing consumer welfare or utility. Perhaps a more reasonable objective and one that recognizes the scarce resource issue is to have an objective of maximizing welfare which means the sum of consumer and producer surplus. Such an objective function recognizes the separate roles of customers and suppliers and the trade-off of increasing costs and increasing quality. Using this approach we should be able to say something about optimal market size, vehicles per market and optimal vehicle size. Below we develop the beginnings of an algorithm that can be used to measure the contribution of AVL to both passenger (customer) welfare and supplier efficiency. The development can be considered in four stages. In the simplest case demand is fixed spatially and market size is fixed. For a given geographic area (so many city blocks) and for a given fixed known demand that is distributed in some way around this geographic area, and for a homogenous type of demand solve for the optimal routing and scheduling according to some objective function. The objective function may be economic welfare (sum of consumer plus producer surplus), to maximize consumer welfare, to minimize costs. All are possible candidates