Take a deep dive into Smart Vehicle Architecture with Aptiv Connected Services, whose Chief Scientist Emilio Frazzoli will be speaking at the IAA Conference.
The historical vehicle architecture approach no longer works – it can’t support the growth in content and complexity the automotive industry is facing today. More safety features and the integration of AI techniques in vehicles mean more computing power, more data and more power distribution than ever before. And as the car becomes a supercomputer, with more and more features and connectivity, its architecture or foundation needs to change radically.
This is where Aptiv’s expertise in smart architectures comes in. Aptiv’s Smart Vehicle Architecture incorporates the full suite of its technology and brings it together through its systems integration expertise. Aptiv is unique in that it can integrate whole systems through Smart Vehicle Architectures, which enable all the electrification, active safety, automation, and connectivity, driving this new mobility.
Aptiv is convinced that Level Four and Level Five autonomous vehicles will experience an architectural break; therefore, the design and production processes must be reimagined. Carmakers have completely innovated what a car can do – without really revolutionizing the way it’s made. The modular way in which features have typically been added to a vehicle has led to an inflexible and unaffordable production process. As advanced driver assist systems evolve from semi-autonomous to fully autonomous, a similar evolution must happen on the production site.
Smart Vehicle Architecture (SVA) is how Aptiv describes the future of the car. While there are a variety of ingredients – electric, active safety, automation and connectivity systems to name a few that make up a car’s architectural ecosystem – Aptiv envisioned SVA as a unified backbone that consists of three critical systems: compute, network and power. And if you follow the systems and circuitry from sensor to cloud – that’s essentially Smart Vehicle Architecture.
SVA reimagines the way Aptiv architects a car, creating the foundation for the software-defined vehicle of the future, while also enabling the flexibility that customers value. Aptiv creates customized architectural solutions that speak to the unique parameters of the problem. What's important is that Aptiv is making the best decisions around architecture to exceed expectations when it comes to flexibility and safety.
It should come as no surprise to read this: cars should be safe. And for years, cars have been designed with safety first. But – what happens to safety when cars themselves are reinvented?
While Aptiv was reimagining ways to create vehicle architecture, it was simultaneously reimagining the safety systems that would keep drivers safe should one of the architectural components malfunction or fail. Aptiv’s team developed a three-layer, fail-operational design which embeds resilience in all three layers (compute, network, power). This embedded resilience means that, in the face of a partial or total breakdown of a system, the vehicle is still able to come to a safe stop.
The key to safe autonomous cars is artificial intelligence (AI). AI is basically the idea that a machine can learn, think and behave like a human. A good example is autonomous driving where your car is literally programmed to react like you would. It’s driven by software and integrated into the whole system.
Aptiv uses AI techniques, specifically neural networks, to allow autonomous vehicles behave like we humans do. It weighs the behavior of the objects around it and balance that against its goal, which is to move into the left lane. This allows the car to decide if it needs to be more aggressive or more conservative in achieving its goal.
To make this possible Aptiv uses a planning platform that relies on pre-determined, generalized rules for basic safe operations and uses artificial intelligence to solve for the optimal path. That means that instructions have been coded into an algorithm, or set of rules the car follows, thereby by creating a vehicle capable of making decisions using AI.
To make a vehicle react safely in a real-world environment algorithms must be “trained” to recognize what’s around the vehicle. Machine learning relies on what called a “neural net” – so-called because it’s designed to behave like the human brain – lives onboard the vehicle and classifies objects in real-time. From here, the vehicle can “follow” specific rules. It’s very complicated stuff and that’s part of the reason why a hybrid approach, which incorporates AI with a machine learning approach, helps autonomous vehicles “drive” more like a human. In certain cases, such as a red light or when another vehicle is stopped in the roadway, it is important the vehicle always stops. But in others, such as when a plastic bag is blowing across the street; it is better for the vehicle to recognize the object is not an obstruction, and it’s safe to continue forward.
Aptiv believes that the power of new mobility – moving safer, greener and more connected – can change the world. And the company knows how to get it done. Aptiv is delivering the software capabilities, advanced computing platforms and networking architecture that makes mobility work. Aptiv is a global technology company that develops solutions enabling the future of mobility. Headquartered in Dublin, Aptiv has approximately 150,000 employees and operates 14 technical centers, as well as manufacturing sites and customer support centers in 45 countries. Visit aptiv.com.
Emilio Frazzoli, Chief Scientist of Aptiv Autonomous Mobility
Emilio Frazzoli is a driving force in developing planning and control algorithms for the safe and reliable operation of autonomous vehicles in real-world environments. In 2013, Frazzoli co-founded leading autonomous vehicle software company, nuTonomy, which was acquired by Aptiv in 2017.
Frazzoli is the pioneering inventor of his Rapidly-exploring Random Trees (RRT) algorithm, which is considered the state-of-the-art in motion planning. He is a world-renowned expert in robotics, fleet management, and autonomous systems. In acknowledgement of his seminal work, Frazzoli has received numerous awards, including the coveted IEEE George S. Axelby Award in 2015 and the IEEE Kiyo Tomiyasu Award in 2017. He has published more than 200 papers in the fields of robotics, autonomous vehicles, and drones. A former professor at MIT, Frazzoli directed the research group that put the first autonomous vehicles on the road in Singapore.
He will be speaking about Autonomous Vehicles: Why "rule breaking" matters on September 12 at 3pm in the Great Auditorium.