Development of safety of the intended functionality from insufficiency of perception and decision making

Introduction : Safety of the Intended Functionality (SOTIF)

SOTIF refers to the safety measures and considerations taken to ensure that a system performs its intended functions without causing harm, even in the presence of foreseeable misuse or unexpected scenarios.
More clearly, the SOTIF is aiming at minimizing the set of unknown and hazardous scenarios. It involves identifying potential hazards, assessing risks, and implementing safeguards to mitigate those risks, thus ensuring the system operates safely under all intended conditions.

Workflow of scenario generation with SOTIF principles

The Vehicle Intelligence and Control Laboratory is conducting a project with the goal of building a system and service that can automate SOTIF testing based on various use cases.
To this end, both simulator-based virtual data and real-world road data are collected to construct a diverse set of scenarios, which are then systematically stored in a centralized scenario database. These stored scenarios are analyzed to extract suitable cases for testing and are utilized in the validation of autonomous driving algorithms.

Application : SOTIF data service (Scenario.ai)

SCENARIO.AI can accelerate the process of developing and validating ADAS systems.
You can build a database to store a large number of scenarios and extract the scenarios you need for AV assessment via queries. You can also analyze the fail scenarios that come out of evaluating AV and use them to develop AV.

• Dataset safety analysis: Analysis of the collected dataset and the training set and test set required to develop the AV.

• Scenario generation: Generated scenarios have to model the overall ODD and foreseeable misuse scenarios and contribute to minimize the set of (unknown) hazardous scenarios

• Database with scenario data: useful for scaling scenario datasets.