Implementability of Artificial Intelligence and Machine Learning Approaches to Vehicle Homologation Tests
Author :
Abstract
Regulations governing the homologation process for regular and/or public transport vehicles require numerous tests to be carried out during the production phase, and it is extremely important that these tests are evaluated accurately, quickly, and at low cost. This article explores the application of Artificial Intelligence (AI) and Machine Learning (ML) techniques to the regulatory testing processes of MAN Türkiye A.Ş., a manufacturer of public transportation vehicles, within the context of current literature, with the aim of improving the speed and cost-effectiveness of these tests. The research examines the most prominent current applications related to the homologation process and includes approaches utilizing AI/ML approaches as well as numerical/statistical analysis for relevant regulations. A key challenge in current homologation processes is integrating information-based system components into traditionally rigid automotive verification systems. In other words, bridging the gaps between Information and Communication Technologies (ICT) and AI/ML in homologation processes. The merging of AI and ML techniques into the current homologation and self-assessment processes applied to passenger vehicles will undoubtedly increase the accuracy and speed of these processes. The information obtained showed that integrating AI/ML approaches with database-supported engineering field applications, especially those with testing and ICT infrastructure in which MAN Türkiye A.Ş. is strong, will provide benefits in validation processes.
Keywords
Abstract
Regulations governing the homologation process for regular and/or public transport vehicles require numerous tests to be carried out during the production phase, and it is extremely important that these tests are evaluated accurately, quickly, and at low cost. This article explores the application of Artificial Intelligence (AI) and Machine Learning (ML) techniques to the regulatory testing processes of MAN Türkiye A.Ş., a manufacturer of public transportation vehicles, within the context of current literature, with the aim of improving the speed and cost-effectiveness of these tests. The research examines the most prominent current applications related to the homologation process and includes approaches utilizing AI/ML approaches as well as numerical/statistical analysis for relevant regulations. A key challenge in current homologation processes is integrating information-based system components into traditionally rigid automotive verification systems. In other words, bridging the gaps between Information and Communication Technologies (ICT) and AI/ML in homologation processes. The merging of AI and ML techniques into the current homologation and self-assessment processes applied to passenger vehicles will undoubtedly increase the accuracy and speed of these processes. The information obtained showed that integrating AI/ML approaches with database-supported engineering field applications, especially those with testing and ICT infrastructure in which MAN Türkiye A.Ş. is strong, will provide benefits in validation processes.
