The last mile problem is crucially important for big cities. To solve it you need to find the ideal balance between different criteria: the mobility of people, comfort, unique characteristics of every person, the profit of transport system and other criteria in more complex models. To solve this problem for dynamic (not fixed) routes part, we offer to use the ontological description of citizens and transport system for matching and multi-agent technologies to plan different types of transport. To prove the idea a prototype of an intelligent multiagent transport system matcher was created. Based on a simple model of Person, Transport and Infrastructure description we proved the feasibility of the approach. In this paper for the first time, we present the Smart Transport System as a personal service provided for citizens. The main idea of the work is to take a step towards the new paradigm where not people adjust their plans for transport but smart transport provides personal service to every citizen according to their needs.
A method of calculating the delivery plan with regards to coordination with a storehouse picker’s schedule in the consideration of intra-city food delivery has been analyzed and is being proposed. In order to assemble a client’s order, storehouse pickers are employed, who, upon notification, collect the necessary items for packaging and delivery. The picker’s assembly schedule should be consistent with couriers’ intended schedule of travel around the city; delivering packages to clients, according to the client’s preferred time windows. It is necessary to
calculate a coordinated and conditionally optimal plan for storehouse pickers on the one hand, and for courier drivers on the other, taking into account their mutual interrelationship. At the same time, it is required to be able to pro-vide a desired time window for delivery of the order to the client. These calculations result in minimizing the courier’s route, taking into account the forecast of traffic congestion at certain times of the day, adaptively redistributing the assembly of packages (order filling) to other stores, if all pickers in the current location are overloaded.
The special VRP problem of transport resources allocation for freight transportation companies that deliver cargo via FTL business model was considered. It was admitted that each real freight transportation company operates in a real time and needs to react on incoming events adaptively reallocate available resources. For this purposes multi agent systems are well proved and used in many modern freight companies. But it was admitted that there is a possibility to improve a quality optimization level by using the stable time period during night hours when no new events come into the system and there is a time for using classic optimization approach in special subproblem of finding the initial allocation plan. By using expert human real logistic scheduling knowledge for a long time period the essential set of limitations to this initial allocation plan problem was defined. The problem was formalized similar to the classical assignment problem of linear programming. Acyclic and cyclic cases of the problem were considered. It was shown that the acyclic case of the problem could be reduced to the assignment problem easily but for the cyclic case it requires to exclude important resource to order matching condition. Finding the exact solution of the initial plan problem was proposed by using the Hungarian method, which is well proved exact method. It was also shown that this method couldn’t be applied in case of real time scheduling, because even in static cyclic case it is impossible to support resource to order matching condition for next future orders, but it can be applied as an addition to the multi agent approach.
The use of multi-agent platform for real-time adaptive scheduling of trucks is considered. The schedule in such system is formed dynamically by balancing the interests of order and resource agents. The system doesn’t stop or restart to rebuild the plan of mobile resources in response to upcoming events but finds out conflicts and adaptively re-schedule demand-resource links in plans when required. Different organizational models of cargo transportation for truck companies having own fleet are analyzed based on simulation of statistically representative flows of orders. Models include the rigid ones, where trucks return back to their garage after each trip, and more flexible, where trucks wait for new orders at the unloading positions, where trucks can be late but pay a penalty for this, and finally where orders can be adaptively rescheduled ’on the fly’ in real time and the schedule of each truck can change individually during orders execution. Results of simulations of trucks profit depending on time period are presented for each model. These results show measurable benefits of using the multi-agent systems with real-time decision making — up to 40-60% comparing with rigid models. The profit dependencies on the number of trucks are also built and analyzed. The results show that using adaptive scheduling in real time it is possible to execute the same number of orders with less trucks (up to 20%).
The special real-time problem of transport resources allocation for freight transportation companies that deliver cargo via FTL business model was considered. Each freight transportation company should react on incoming events adaptively reallocate available resources. For this purposes multi agent systems are well proved and used in many modern freight companies. But it was admitted that there is a possibility to improve a quality optimization level by using classic optimization approach in the special initial allocation subproblem. By using expert human real logistic scheduling knowledge for a long time period the essential set of limitations to this initial allocation plan problem was defined. The problem was formalized similar to the classical assignment problem of linear programming. Acyclic and cyclic cases of the problem were considered. It was shown that the acyclic case of the problem could be reduced to the assignment problem easily but for the cyclic case it requires to exclude important resource to order matching condition. Finding the exact solution of the initial plan problem was proposed by using the Hungarian method, which is well proved exact method. It was also shown that this method couldn’t be applied in case of real time optimization, because even in static cyclic case of the problem it is impossible to support resource to order matching condition for next future orders, but it can be applied as an addition to the multi agent approach.
The application of multi-agent platform for real-time adaptive scheduling of trucks is considered. In case of unpredictable events the system works adaptively and doesn’t stop to restart the plan from the beginning. Different models of cargo transportation for truck companies having own fleet are analysed. The results show that using adaptive scheduling in real time it is possible to create significantly more profitable schedules (up to 40-60% compared with rigid models) and save a number of trucks (up to 20%) for the same amount of orders.