Automated vehicles development has accelerated, and this change is fueled by technology, consumer behavior, and new business models. We have seen how this new technology has influenced (or is just starting to impact) numerous industries, including insurance, oil, and the entire mobility sector.
So what else is going to change?
Autonomy will change driving and transportation as we know it. Believe it or not, tires will be changing significantly with these new developments. Automated driving, electrification and shared mobility are shifting the requirements on tire performance. This then affects how tires are being tested and modeled to support the development of new tires and vehicles. Let’s explore how this monumental autonomous movement will affect future tire testing and modeling.
When you are the driver, your main focus is on driving the car from A to B, whilst a passenger can make use of the travel time to do other things. In current cars, it is important for the driver to feel in control of the car and connected to the road, not only for the pleasure of driving, but also for safety reasons. On self-driving vehicles, the passengers’ hands and focus are free to check emails, read, or scroll through social media. Steering feedback is no longer important, however the rider comfort and safety are. This adds emphasis for tire manufacturers to provide not only a traditionally well-performing tire, but also one that provides a smooth ride for passengers. Tire vibration and noise will be more noticeable than ever without a combustion engine, so there will be an increased focus on reducing NVH in future tire testing.
How will tires influence autonomous vehicles?
Another area where tire models will play a key role will be in the trajectory planning and vehicle control during emergency maneuvers. The image below shows an example control module for a self-driving vehicle. The perception module takes all the information from the available sensors and builds up an image of what is around the car. This image is fed to the planning module that will be responsible for planning the next action to take and the control module will be responsible for ensuring that the car achieves the requested action.
In any situation, the autonomous system must decide where and how to drive next. This includes normal driving, changing lanes, emergency maneuvers, etc. Most of the time, tires will be operating in the linear range, in which a change in force is proportional to the change in slip. Current vehicles with self-driving capabilities can handle dynamic situations up to a certain level but will ask the driver to take control when the car deems the situation too complicated.
However, for a vehicle to be classified as a level 4 or 5, which do not require a human driver, they must be fully able to handle all situations, including emergency maneuvers. In highly dynamic maneuvers, the system must be able to control the vehicle at high levels of combined slip to maximize the grip potential of the tires. This will require autonomous vehicles to be as good or better than human drivers, by using more data and coming to better decisions faster.
In current cars, a simple tire model works fine. However, as self-driving cars become more advanced and take more responsibilities, the tire models used in the system need to evolve to help the planning and control modules achieve more accurate control of the vehicle.
New tire sensor technology can be used as sensors by providing valuable input to the self-driving system about current tire and road conditions. This, in turn, can then significantly affect prediction and trajectory planning, as well as overall vehicle control. The system is evaluating thousands of possible trajectories in order to determine the best path in any situation. For the system to be as precise as possible to make the best-informed decision, it will need highly accurate tire models derived from previous testing and current conditions. This becomes extremely important in emergencies, where there may only be a few trajectories to evade danger.
How might this play out on the road?
Let’s look at an example.
The graph below shows a car going down a straight road and it suddenly encounters an obstacle. The lines represent all the possible trajectories calculated by the system in the initial iteration with a linear tire model.
In the example, out of 1000 calculations, about half of them would cause a collision (red), the other half would drive the car out of the road (orange) and only a few trajectories would avoid the obstacle (black).
The graph below shows one of the successful trajectories (black line) from the previous example recalculated with a full Pacejka tire model (red line). As seen below, the vehicle would not achieve the predicted trajectory with the pre-calculated steering input, and the vehicle would not avoid the obstacle. Of course, the control module would continuously correct this deviation, but in such a short time, the corrections might be too late to avoid the obstacle.
In this case, a more accurate tire model would have helped improve trajectory planning earlier on, and would have helped to significantly increase the likelihood to avoid the obstacle. Combining this with the situational analysis of external factors on tire performance can help to create a safer and more reliable self-driving system.
New tire testing and modeling methods will be needed to address the specific requirements of autonomous vehicles. This is an area where integration and partnership between vehicle and tire manufacturers and testing/modeling companies is essential to drive innovation and to share knowledge. Calspan welcomes further discussion and ideas on these topics! If you’d like to connect, reach out to Mateo Gladstone or contact another Calspan expert here.
Using the world’s most powerful flat track tire testing machine, combined with more than 40 years of tire research and testing experience, Calspan’s talented engineers support tire and vehicle manufacturers worldwide with R&D projects and bringing new innovations to market. Measuring tires can be a challenging task, but, thanks to proven quality procedures, measurement quality and repeatability can be ensured, making subsequent analysis and modeling more precise and efficient. We have continuously honed our abilities to stay current with trends and new developments in the market. Get in touch with Calspan to see how we can help with your next innovation.