Based on ARTIFICIAL INTELLIGENCE and the secure FRAMEWORK, EYYES offers with its product suite ROADEYE solutions for the SENSORFUSION of sensor technology for road traffic infrastructure.
Due to the long-standing know-how of EYYES experts with event detection and sensor systems in the field of roads, existing control systems can be retrofitted via standardized interfaces and sensor fusion solutions can be integrated into new plants. This increases the relevance of the messages. The number of false alarms can be significantly reduced.
TUNNEL SENSOR FUSION.
Many different sensor systems are used to monitor road tunnels, which are recorded and processed by, among others, process control and SCADA systems and displayed and operated in control rooms and traffic control centres.
Each sensor system functions within the set requirements and complies with a defined detection or misdetection rate, also called sensitivity and specificity.
However, since the individual systems are disconnected, i.e. analyse the respective event types independently of each other, the false alarm rates add up, for example statistically, with each additional subsystem and sensor.
Typically, such sensor systems are:
Fire alarm systems and smoke detectors
Video detectors for the detection of lost cargo,
wrong-way drivers, fire/smoke or, for example,
Audio detectors for the detection of exceptional
Thermal or ground radar systems
Environmental and air measurement systems such
as weather sensors, turbidity meters, CO detectors
Loop systems ‒ optical height controls, induction loops
Contact-based sensors such as escape door contacts, emergency telephone niches or the removal of fire extinguishers
AVIFUSION as middleware between sensor and process control systems merges and processes the incoming messages. Individual sensor sources are identified, the propagation and development of the frequency of messages are tracked and evaluated, thus effectively filtering the information from sensor systems, suppressing multiple messages and significantly reducing the number of false alarms. New and highly prioritized events are recognized as such and intensified accordingly.
This concentration on actual messages substantially increases system acceptance and usability of the overall system.
In case of damage, messages from many and different sensors result in an unclear picture of the situation, the starting point and the cause cannot be clearly determined.
The source, spreading and development of, for example a fire cannot be clearly traced due to the large number of sensor messages ‒ AVIFUSION not only merges the messages, but also analyses all sensor data. The resulting situation picture represents a much improved view of the damage event or incident. The starting point of the event can be determined and visualised as well as the triggering event itself.
In addition, HISTORICAL DATA are included in the evaluation, which, for example,
include hotspots with accident clusters or their typical progression in the analysis of situation.
TRAFFIC FLOW ANALYSIS.
Based on the framework for the AVI REALTIME AI ENGINE AVIFLOW allows vehicles to register when entering a specific area, such as a road section, and to deregister accordingly when exiting.
AVIFLOW records various vehicle parameters such as vehicle class, type, brand or registration number. Based on these data, third-party systems can count and categorise vehicles to create a traffic flow analysis. Therefore, distance and time-dependent analyses can be created.
Another application is the surveillance of environmental zones or sections with weight restrictions ‒ raw data are transmitted via flexible interfaces to the authorities for further processing and evaluation.
Due to its flexibility, the sensor technology of the system can be used for a large number of analyses and is designed with mechanisms that comply with data protection requirements.