With StreetWise, TNO has developed a methodology to test and validate complex automation and driver assistance systems. A large number of traffic scenarios and test cases have been distilled from ‘real-world’ vehicle data that are suitable as scenario simulation data. Torc Robotics uses Streetwise to prepare self-driving trucks for all possible traffic situations that may occur between two hubs. The truck driving data is analyzed and categorized into scenarios using StreetWise. This allows Torc to collect a large number of relevant scenarios for extensive virtual testing of autonomous technology according to the latest international safety guidelines and protocols.
Pioneers in automated driving
The American Torc Robotics from Blacksburg, Virginia, is one of the pioneers in the world of automated driving. The independent subsidiary of Daimler Trucks specializes in software for self-driving vehicles and is currently focused on commercializing autonomous long-haul trucks in the US. Axel Gern, Senior Vice President of Engineering and Managing Director of Torc Europe: “We have 17 years of experience in safety systems for self-driving vehicles, and our technology is designed with safety in mind. We try to emulate the driving behavior of the safest, most experienced drivers. We use TNO’s StreetWise methodology as the basis for our safety validation. We also share the same views on road safety. That is why TNO understands our needs and they support us in implementing and scaling up this methodology.”
Collective traffic scenarios
For safety testing, the industry uses traffic situations that are recorded in a parameterized model, as defined for example in the “Pegasus project”. The core of the StreetWise methodology consists of a database in which all these traffic situations are stored. The scenarios are based on real driving data from different fleets of vehicles. They form valuable input for smart simulation software with which you can calculate the residual risks of autonomous driving. TNO’s methodology also allows multiple partners to build a collective scenario database together, without having to exchange sensitive data. That saves time and costs. Instead of running expensive testing programs or collecting data in each region, they can scale much faster to a complete scenario database for different cities, countries and continents.
New traffic insights
TNO and Torc are working together to implement the StreetWise software in Torc’s fleet operation. They use the software to process the daily flow of data from the test fleet and this gives the engineers new insights. Olaf Op den Camp, specialist in safety assessment at TNO: “Autonomous transport plays an essential role in TNO’s mission to make traffic and transport more efficient, sustainable and safer. The new security challenges require legislation and standardization within the industry. We are very pleased that Torc is the first party to adopt our methodology, bringing security validation adoption and standardization one step closer.”