Beneficiary name: Csiha Innovation and Technology Plc.
Project title: Artificial intelligence-based vehicle traffic load detection sensor and decentralized
decision support system for transport nodes implemented with a decision network
Project identification number: 2020-1.1.2.-PIACI KFI-2020-00018
The contracted amount of the project: HU 351 243 305
The agreed grant amount: HU 232 505 228
Support intensity: 66.19%
Project completion date: February 29, 2024.
Project content:
The project aims to design and create a hybrid sensor group and control unit based on advanced technologies,
suitable for vehicle and terrain detection for traffic management purposes.
In the traffic management systems currently in use, traffic intersections usually receive their traffic light schedules from a
central location, which operate according to a predefined algorithm. As a result, they are unable to react to certain situations, or only do so more slowly and with greater difficulty. With increasing urbanisation and the dramatic growth rate of transport means, there is a growing need for an intelligent, traffic-
decentralised, autonomous decision making system that is optimised for traffic flow. This need has already been identified in several cities in Hungary, but is even more pronounced at international level, especially in cities of major commercial and logistical importance.
In this project, we are implementing an intelligent traffic management system,
which can reduce the difficulties of everyday transport and parking. Our company
extensive practical and research experience in vehicle detection and artificial intelligence
development, so we are creating a system based on a self-designed
a set of vehicle detection sensors and a control system with complementary artificial intelligence.
The project will develop a decentralised decision maker to collect sensor data,
a self-learning and in-situ analysis artificial intelligence and control system, which will also
will be implemented under the project. The sensor arrays will be installed at traffic intersections,
where artificial intelligence enables them to act as autonomous decision-making units, while at the same time
will also be integrated in a large network. Standalone groups are able to take into account the network
decisions of other operating units and thus coordinate their operations. The self-learning, autonomous decision maker
units can operate completely off-line, no need to connect to the central server and other cells
to be constantly connected in order to make it work.
The artificial intelligence that we develop continuously analyses the data after installation and
processes the passing vehicles to get a realistic picture of the traffic at a given junction. The resulting
data that allows a node to autonomously assess and manage traffic on the given traffic
according to the situation. The AI will guide you after processing the incoming data,
to divert traffic so that traffic can flow as quickly and smoothly as possible. Meanwhile
no need for external intervention or a remote, expensive server, because the system is completely
self-learning and automatic, and performs calculations cost-effectively on-site, installing and
maintenance is cheap: no need to demolish pavement, it can be installed in a built-up junction.
The efficiency of the design lies in the fact that the sensor nodes are coordinated
work, which speeds up everyday transport: time
save money, reduce the risk of accidents and help protect the environment by lowering emissions. This special decision-making network and the
measuring vehicle traffic at intersections can therefore be used in many other industrial areas. For parking systems,
in underground car parks, airports, markets, where sensors can be used to provide the most effective control
and use of space.
As part of the research and development, several nodes will be deployed to provide the appropriate testing
can automatically control traffic after a period of time, and can react to unusual but
repetitive traffic morals, situations. The service provided by our product offers a high level of safety
solutions, while minimising the risk of negative decisions due to the human factor
the occurrence of. There are already various vehicle detection technologies on the market, but they
are typically cumbersome and/or expensive to install, and are typically on-line and stand-alone
are not suitable for decision-making. In contrast, the technology offered by the system we are developing
is unique in the market, can be cost-effectively installed in existing nodes, and offers a range of multi
for the operator (city, car park and event centre management) offers fewer and cheaper elements
package, which significantly reduces development and operating costs, while at the same time
decision making, it represents a significant innovation in the market.