Solution concepts

The solution proposed by the PEGASI team assumes the collection of data on people and vehicles moving on the premises of the Main Campus of the Warsaw University of Technology for such purposes as design and space management using edge computing. Great emphasis is placed on data anonymization. Space management based on ongoing information includes, among others, intelligent lighting control and signaling management in order to change the way space is used and at the same time increasing safety. In general future traffic management will take into account objects pedestrians, cars as well as small vehicles including bicycles and scooters and areas of their crossing.

The project assumes the possibility of later use of the experience gained in other urbanized areas, hence one of the basic assumptions was the scalability of the solution, minimization of costs and the use of technical infrastructure already partly present in urban areas, including municipal monitoring.

In order to collect data on users, it was decided to first collect data on the flows of people and vehicles on individual paths and analyze their propagation from nodal points - the intersection of paths described hereinafter as edges. Hence, the analyzed space was divided into nodal points and edges visible in Fig. 1. As the next stage of tracking pedestrians and vehicles, the possibility of tracking the paths of moving objects through the analyzed area, taking into account the speed of their movement and stopping places, taking into account the principle of data anonymization. The collected data was supposed to allow for obtaining information about pedestrians and vehicles, distinguishing the time intervals of use on particular days of the week.

Ultimately, it was decided to use nodal points from CCTV cameras for the analysis and, additionally, to install beam intersection sensors to analyze traffic intensity at individual edges. The location of the cameras and measurement points is shown in Fig. 2. 

Work on data collection and the possibility of using individual sensors hardware and software tests in laboratory and field conditions. Two basic data collection methods were tested:

• CCTVs sensors

• Beam intersection sensors