The issues of forecasting the load of vehicles, route networks, formation of collection centers for cargo flows, including dangerous goods, in territorial planning systems determine the need for transition from deterministic models to dynamic models with the possibility of taking into account the influence of the external environment. Existing information transportation systems in the sphere of work with dangerous goods are based on deterministic models and allow to study the problems of forecasting changes to a limited extent. Models of routing networks construction on cargo vehicles are limitedly applicable when there is a task of dangerous goods transportation taking into account the presence of a distributed system of accumulation centers both in the city and, for example, within a large transport hub, seaport with different terminals. The paper presents routing models adapted to the conditions of dangerous goods transportation. Determining transport efficiency and the main costs of transport enterprises is the scheme of traffic flows, fixed in the territorial scheme. The task of the study is to develop a separate automation toolkit for decision making on the number and type of specialized transport (depending on the load capacity) to work with dangerous goods and large-sized objects within the existing flow scheme. On the basis of the research of the group of companies the conclusion is presented that today the task of choosing the optimal amount of transport by carriers is solved, as a rule, not systematically, but by operative construction of a set of routes, and does not involve primary analysis of the territory and parameters of accumulation sites. The presented study considers a new approach to the analysis of the system of transportation of hazardous waste, associated large cargoes, in terms of assessing the necessary list of universal structured input data, which are primary principled and characterize the complexity of work in the selected facility, terminal. To solve the problem of forecasting the load of vehicles is offered developed a new automation subprogram based on the organization of client-server access to data, and implemented in the internal programming language “1C. Management of motor transport” (version ММT 8.2). The presented automation subprogram provides data connectivity on volumes of dangerous goods cargo flows, cargo flow collection centers, available data on vehicles and planned data on removal schedules. Work of the subprogram provides integration of work with other databases of transport companies and allows to carry out modeling of various variants of placement of centers of collection of dangerous goods, centers of cargo flows with high accuracy. The obtained results allow to determine the required number of vehicles, their load taking into account different types and technical capabilities, as well as to determine the required number of vehicle trips in the serviced area taking into account the dynamic influence of the external environment. The presented methodology of working with data, performed in a separate automation subprogram, presented routing models, adapted to the transportation of dangerous goods, can be applied to various transport systems, including sea cargo ports and terminals, which determines the universality of the presented approach.