REDA: A New Methodology to Validate Sensor Systems for Person Detection under Variable Environmental Conditions
Authors:
- Christian MeltebrinkCMChristian Meltebrink
- Magnus KomeskerMKMagnus Komesker
- Carolina KelschCKCarolina Kelsch
- Daniel KönigDKDaniel König
- Mario JenzMJMario Jenz
- Marvin StrotdreschMSMarvin Strotdresch
- Benjamin WegmannBWBenjamin Wegmann
- Cornelia WeltzienCWCornelia Weltzien
- Arno RuckelshausenARArno Ruckelshausen
Open Access
Abstract
Perception of the environment by sensor systems in variable environmental conditions is very complex due to the interference influences. In the field of autonomous machines or autonomous vehicles, environmental conditions play a decisive role in safe person detection. A uniform test and validation method can support the manufacturers of sensor systems during development and simultaneously provide proof of functionality. The authors have developed a concept of a novel test method, “REDA”, for this purpose. In this article, the concept is applied and measurement data are presented. The results show the versatile potential of this test method, through the manifold interpretation options of the measurement data. Using this method, the strengths and weaknesses of sensor systems have been identified with an unprecedented level of detail, flexibility, and variance to test and compare the detection capability of sensor systems. The comparison was possible regardless of the measuring principle of the sensor system used. Sensor systems have been tested and compared with each other with regard to the influence of environmental conditions themselves. The first results presented highlight the potential of the new test method. For future applications, the test method offers possibilities to test and compare manifold sensing principles, sensor system parameters, or evaluation algorithms, including, e.g., artificial intelligence.Keywords
Funding Information
- Bundesministerium für Forschung, Technologie und Raumfahrt (13FH642IX6)
- B. Strautmann & Söhne GmbH u. Co. KG (13FH642IX6)
This publication has 12 references indexed in Scilit:
- Deep Learning in Perception of Autonomous Vehicles2022
- An overview of sensors in Autonomous VehiclesProcedia Computer Science, 2022
- Evaluation of Warning Methods for Remotely Supervised Autonomous Agricultural MachinesJournal of Agricultural Safety and Health, 2022
- An Approach for Evaluating the Completeness of the Test Scenarios for the Vehicle Environmental Perception-Based Systems2021
- Multilayered review of safety approaches for machine learning-based systems in the days of AIJournal of Systems and Software, 2021
- Concept and Realization of a Novel Test Method Using a Dynamic Test Stand for Detecting Persons by Sensor Systems on Autonomous Agricultural RoboticsSensors, 2021
- Vision-Based Safety-Related Sensors in Low Visibility by FogSensors, 2020
- A Systematic Review of Perception System and Simulators for Autonomous Vehicles ResearchSensors, 2019
- Test methodology for rain influence on automotive surround sensors2016
- The Third Criterion: Compactness as a Procedural Safeguard Against Partisan GerrymanderingSSRN Electronic Journal, 1991