The Abu Dhabi Autonomous Racing League (A2RL) is an exciting new concept in the realm of motorsport, where the world’s brightest minds compete to develop, and race, incredible racecars using solely Artificial Intelligence (AI). Research and education institutions from across the globe will compete with one another, battling on track for a stake in the US $2.25 million prize purse. Ten cars will be released for the inaugural race in 2024. Instead of drivers, the ‘heroes’ of the sport will be the engineers, scientists, programmers, and technicians. A2RL is the brainchild of ASPIRE, the technology transition pillar of Abu Dhabi’s Advanced Technology and Research Council. The competing teams involved in the inaugural season include:
• Beijing Institute of Technology and Khalifa University (China and UAE) • Code19 Racing and Indiana University (USA) • Constructor Group (Switzerland & Singapore)
• Hungarian Mobility Development Agency (Hungary)
• Kinetiz & Nanyang Technological University (Singapore)
• Politecnico di Milano (Italy)
• Technical University of Munich (Germany)
• University of Modena and Reggio Emilia (Italy)
What are the main aims of the series?
The motivation behind A2RL is to build an ecosystem around autonomous mobility. This will accelerate the development and effectiveness of autonomous systems for deployment in road cars. Many drivers are showing a lack of enthusiasm towards ADAS 3 components such as Lane Assist, which they find disconcerting. The competitive environment of motorsport can demonstrate the reliability and safety of these systems thus enhancing consumer confidence and de-risking OEM investment. Innovative race formats will be used to push the frontiers of current technology and deliver entertainment and insight to audiences at the track and online. STEM events will form an essential component of A2RL. In parallel with the inaugural race a STEM competition for students is being organized where they will code, and race scaled down versions of autonomous vehicles. A central motivation in establishing A2RL was to position Abu Dhabi as a hub for innovation in autonomous mobility. We are partnering with world class suppliers in adapting and developing the Dallara Super Formula car for autonomous racing; attracting leading teams from across the globe to come and compete in Abu Dhabi; and creating a platform for youth to take active involvement, demonstrating the attractiveness of a career in engineering and science and making the consumers of the future more informed. All together this will ensure that the innovation we are promoting will stick locally.
How has the SF23 racing car been adapted?
While it may look fairly similar from a distance, the modifications done to the SF23 racecars used in A2RL are extensive. All race cars in A2RL use a drive-by-wire system that replicates human inputs. This system employs actuators for the steering, braking, and gear shifting, eliminating the need for a physical driver. These actuators are all controlled by the onboard computer. The next major modification is the vast array of sensor technologies that will allow the machine learning algorithms to ‘see’ and develop situational awareness – including radar, lidar, cameras, and GPS. These will be supported by a number of different control modules. Each will supply the driving AI with a continuous feed of data, which can be used to determine the best course of action. Essentially, the AI is alone on the track, with the only outside input being an emergency kill switch. All cars will be supplied with a basic level of autonomous performance to get the teams started. It will then be up to them to further optimise the systems and compete with one another on track. They will also take part in two weeks of training at Yas Marina ahead of the annual race to fine-tune their algorithms. This will result in highly competitive racing with immense complexity and excitement – a never-before-seen spectacle.
How is AI being utilised?
AI is the only way that the vehicles can be controlled on track. The development of the algorithms that will interpret data and make ‘decisions’ on the track while racing is entirely up to the teams, beyond the basic functionality that we will provide. The software itself is comprised of various modules that will be used to understand what the vehicle is doing and what it should do next. When you think about all the intuition and minuscule actions and decisions that drivers make instinctively, there is a lot for these systems to understand and work with. That includes tactile feedback from the environment like grip levels, g loads, and suspension movement to give the AI a complete picture of its dynamic situation. Perhaps most important of all is the planning module, which allows the racecar to prioritise objectives and strategise during the race. As you can imagine, it is extremely complex.
What will the race day experience be like for fans/viewers?
This is an interesting question, as without drivers, it is different to almost every other motorsport event that has ever been held. While the focus is on development and driving innovation in the realm of autonomous driving, we have some exciting ideas to engage audiences both on the track and at home. A lot of this will be about a combined digital and physical experience, which you’ll hear more about as we get closer to the race in April 2024.
What other vehicles may you include in the future?
At the moment, our focus is on the Super Formula SF23 and establishing A2RL as a respected and competitive international motorsport. We will launch an autonomous drone race in 2025 and in the future, we will extend to other platforms such as dune buggies.
What will driverless urban mobility be like by 2030?
It is likely that the majority of production cars will incorporate autonomy at ADAS level 3. Our ambition is to provide a testbed that allows OEMs to use extreme sport. Studies around the world have already demonstrated that driverless vehicles are statistically safer than those controlled by humans. There is, however, a substantial disconnect between public sentiment and trust in autonomous systems and their real-world effectiveness. Part of A2RL’s mission is to address these concerns while also spurring the continual improvement of these systems. When you consider 2030, there will have been at least six seasons of A2RL development, alongside the efforts of automotive OEMs around the world. While it may be too early to give a definitive timeframe, it is clear that driverless urban mobility will be a critical part of the transport landscape over the coming decades.
Source : Engineering magazine
