Congestion detection

Within the framework of a service contract, the development of Szolnok City's automatic traffic counting system with a congestion detection function, taking into account the following, as detailed in the public procurement documents. During the project "Establishment of the automatic traffic counting system of Szolnok City", a real-time traffic measurement system was established in the city's 7 city gates and 10 further traffic junctions. The system provides data on the number of vehicles passing through the respective cross-sections via a web interface in real-time, broken down into 30-second time units, from a total of 79 data collection points. The aim of this development is to expand this system with new, congestion detection functions, based on the automatic traffic counting system of Szolnok City and the data obtained from it. The system to be built is expected to include the establishment and integration of the necessary hardware and software systems with the existing system, as well as the execution of its test runs. The established infrastructure, the right to use the software components, and the generated data must be the property of the contracting authority. In congestion detection, the winning bidder's task is to model traffic for the entirety of the examined road sections using the data from real-time traffic measurements, and to derive the parameters characteristic of congestion, defined below, from the model. Congestion detection must be implemented based on the data obtainable from the existing automatic traffic counting system, based on mathematical modeling, without installing additional sensors. The winning bidder must integrate the software implementing congestion detection into the "automatic traffic counting system of Szolnok City". In congestion detection, the winning bidder is obliged to examine the road sections between junctions C5-C4, C4-C8, and C3-C4 for all lanes and both directions (the location of junctions and sensors is in the technical documentation). For a given examined road section, the winning bidder must determine, for 50m segments of the road section, per lane, at a given point in time, the number of vehicles in the lane segment, the coverage of the lane segment by vehicles (given as a percentage), and the average instantaneous speed of the vehicles in the lane segment. The task includes determining the average transit time for the road sections involved in congestion detection, with respect to the sections between the given traffic junctions. The winning bidder's task is to estimate the relevant statistical parameters necessary for calculating coverage by collecting data. The bidder must present the chosen estimation method in their offer. The winning bidder's task is to determine all the expected parameters of congestion detection with a temporal frequency of 2 minutes, and to save the generated data in a database. Web access to the generated data must be provided – retroactively for 1 year. The congestion parameters must be displayed on a web interface, and the retrospective viewing of this data must be ensured for a period of 1 week on the website. The congestion detection software must have a prediction capability of at least 15 minutes. The accuracy requirement for the congestion detection software is: 10% relative error ((calculated value - measured value) / measured value) for the calculated parameters with respect to real-time data, and 20% relative error for predicted data. The total quantity and technical description will be detailed in the technical documentation provided as part of the public procurement documents.