Version 1
: Received: 25 December 2023 / Approved: 26 December 2023 / Online: 29 January 2024 (12:25:22 CET)
How to cite:
Garikapati, D.; Liu, Y.; Huo, Z. Systematic Selective Limits Application Using Decision Making Engines to Enhance Safety in Highly Automated Vehicles. Preprints2024, 2024012034. https://doi.org/10.20944/preprints202401.2034.v1
Garikapati, D.; Liu, Y.; Huo, Z. Systematic Selective Limits Application Using Decision Making Engines to Enhance Safety in Highly Automated Vehicles. Preprints 2024, 2024012034. https://doi.org/10.20944/preprints202401.2034.v1
Garikapati, D.; Liu, Y.; Huo, Z. Systematic Selective Limits Application Using Decision Making Engines to Enhance Safety in Highly Automated Vehicles. Preprints2024, 2024012034. https://doi.org/10.20944/preprints202401.2034.v1
APA Style
Garikapati, D., Liu, Y., & Huo, Z. (2024). Systematic Selective Limits Application Using Decision Making Engines to Enhance Safety in Highly Automated Vehicles. Preprints. https://doi.org/10.20944/preprints202401.2034.v1
Chicago/Turabian Style
Garikapati, D., Yiting Liu and Zhaoyuan Huo. 2024 "Systematic Selective Limits Application Using Decision Making Engines to Enhance Safety in Highly Automated Vehicles" Preprints. https://doi.org/10.20944/preprints202401.2034.v1
Abstract
Safety limits application has always been a traditional approach to ensure the safe operation of electro-mechanical systems within many industries including automated vehicles, robotics, aerospace, traditional automotive, railways, manufacturing, etc. In all of these applications, control and safety limits are usually hard-coded into the production firmware and are fixed throughout the entire product life cycle. Currently, due to the evolving needs of automated systems like automated vehicles and robots, this traditional approach does not address all the use cases and scenarios to ensure safe operation. Especially for data-driven machine learning applications that constantly evolve and learn over time, it is important to be able to adjust the safety limits application strategy to be more flexible and adaptable based on different Operational Design Domains (ODDs) and scenarios. Our ITSC conference paper ~\cite{4} introduced the concept of a new safety limits application strategy called the Dynamic Control Limits Application (DCLA) strategy that supports the flexible application of diverse limits profiles based on the parameters involved within the dynamic scenario at different layers of the Autonomy software stack.This paper extends the DCLA strategy to derive the complete methodology for safety limits application based on ODD elements, scenario identification and scenario classification using Decision Making Engines. It leverages the layered architecture introduced in the ITSC conference paper to implement the Decision Making (DM) algorithms. Another important extension is the use of cloud infrastructure that is based on the Vehicle-to-Infrastructure (V2I) technology to store the scenarios and the limits mapping to serve as a ground truth and/or a backup mechanism in case of errors or failures associated with the main Decision Making (DM) Engine. There is also a focus on providing a more comprehensive list of scenarios and a custom built experimental dataset to cover the maximum ODD elements, and multiple tables of safety limits to be chosen from, which eventually helps in creating different safety limits application profiles. These distinct safety limits application profiles are based on the scenario parameters that are perceived by the system upon which the Decision Making algorithms are applied or trained. This systematic approach can be used within the industry for any of the future automated vehicles and systems until Level 5 Autonomy.
Keywords
automated vehicles; connected vehicles; V2I; multi-criteria decision making (MCDM); Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS); machine learning; safety limits; cloud-based; multiple safety profiles; operational design domain(ODD)
Subject
Engineering, Control and Systems Engineering
Copyright:
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
