Claytex – Technology In Transport

Transport is undergoing a number of technological changes that have the potential to dramatically affect every aspect of this sector including how the vehicles are powered, controlled, owned and used. An increasing number of vehicle manufacturers are announcing their intention to produce hybrid or electric vehicles and to make them fully autonomous over the next 3-5 years. Ride sharing services such as Uber are also working on autonomous vehicles which could add a new dimension to their services with the potential to disrupt more than just the taxi industry.

These changes present many challenges for the engineers working on the design, development and testing of these vehicles to make sure they work correctly and safely. The complexity of the next generation vehicles will mean it is not possible to carry out comprehensive physical testing of all the vehicle systems in every conceivable scenario. Researchers estimate that to prove with 80 percent confidence that autonomous vehicles are 90 percent safer than human drivers, each vehicle being designed would need to drive 11 billion miles. This also implies that for every hardware and software change the full range of tests would have to be repeated making physical testing impractical.

To address this the manufacturers will have to make extensive use of simulation and virtual testing to validate their new designs. For this to be effective the models used will need to include all the vehicle systems and their different physical domains (mechanical, electrical, thermal, cooling, and control). In addition, the virtual test environments need to provide a realistic environment with the possibility to immerse the vehicle sensors and the driver into the same environment. Claytex, and our partners, have been developing technologies to meet these challenges for many years.

By using virtual testing the scenarios that the car needs to be tested under can be condensed into a much shorter distance. Millions of miles of real-world driving can be condensed into a few hundred miles in the virtual environment.

Some aspects of vehicle development have relied on virtual testing for many years, such as control system validation using hardware-in-the-loop (HiL) test rigs where the actual vehicle control unit is connected to a real-time computer. The computer runs a simple model of the vehicle that can generate the sensor signals the control unit normally receives so that the controller recognises its environment. Typically, HiL rigs are just used to test one controller in isolation and the model on the real-time computer is as simple as possible. However, there are a number of issues with the current approaches that mean they need significant changes to support the testing necessary for the next generation of vehicles.

Making best use of simulation for control system development and testing, demands high quality physical models to simulate the behaviour of the system being developed. These models need to capture the behaviour of the complete system including all mechanical, electrical, thermal and fluid systems. Due to the complexity of the next generation of vehicles with many more interconnected systems, the vehicle system models will have to be created using modelling languages that cover all of these different domains.

One modelling language that has been designed for this purpose is Modelica. Modelica is an open source modelling language that Claytex has been using along with the modelling and simulation tool Dymola. Together this tool and language provide us with a powerful, flexible modelling solution capable of modelling all the vehicle systems and, most importantly, giving us the freedom to easily extend its capabilities to cover new technologies.

A detailed and physics-based, rather than empirical-based vehicle system model is key for control system development and for testing of diagnostics and failure modes, particularly when relying on virtual testing to prove how the product will behave in the real world. Models should also be able to cater for scenarios such as sensor failure due to soiling and other situations representative of use within the real world. The controllers will be required to operate precisely with narrow tolerance bands and detailed physical models will aid their development.

For autonomous vehicles, the environment in which the vehicle will operate needs to be well represented. For this purpose, immersive driving simulator environments like rFpro, already in use by motorsports and passenger car OEMs, can be used for simulating the environment that the vehicle and all its sensors will be operating in. These tools can be coupled to the vehicle physics models and real controllers via HiL rigs.

Using rFpro, a camera sensor can be positioned inside the virtual vehicle with the appropriate field of view, frame rate and then lens distortion effects can be applied to generate an image as it would appear on the camera sensor. The sensor model can then replicate the communication protocol used by the real device and this can be transmitted to the real controller via the HiL test rig. The same can be done for all the vehicle sensors allowing the control systems to be fully immersed into the virtual environment.

The final challenge is the complexity of the control system design. This needs to be well structured not only because of regulations like ISO 26262 but also because the complexity of the systems and how they interact; it would be impossible to keep track of these systems without the use of specialist tools and methodologies. We also believe it is necessary to change the practice of the control system developer also creating a simulation model to test their controller. Often these are either based on test data rather than physics or use the same simplifications and assumptions that led to the design of the control system. This means that the control systems are often designed and tested against models that do not fully represent the real system and then problems are only discovered when the first prototype becomes available.

Claytex has been working on solutions to address all of these problem areas. Using state-of-the-art systems engineering solutions that work seamlessly together we can help address all these challenges.