The impact of technology on the automobile industry will continue to increase with the rise of autonomous vehicles.
Autonomous Vehicles (AV)
Popularly known as self-driving cars, autonomous vehicles are automobiles that can operate and navigate with minimal human input by leveraging various integrated technologies, including but not limited to: artificial intelligence, sensors, big data, IoT connectivity, and cloud computing. While still in its early stages, autonomous vehicle technology is considered to be a disruptive innovation that will rapidly revolutionize the entire transportation industry in the coming years.
From the perspective of design engineers and automotive manufacturers, autonomous vehicles are software-intensive mechatronic systems integrating more than a hundred electronic control units (ECUs) connected through various types of networks. Fully-functional autonomous driving systems will require some of the most complex software implementations that carmakers have ever faced – combining a variety of data feeds (e.g. information from sensors, traffic data from the cloud, data coming from other vehicles or infrastructure), and tying it all into the vehicle's electronic and mechanical components to create a network of onboard systems that all work together reliably without user input or correction.
To address the challenges of autonomous vehicle development going forward, automotive companies must rethink the traditional product lifecycle and adopt an engineering methodology that utilizes agile, model-based development with integrated data flows (i.e. digital threads) and robust software-based simulation capabilities.
Autonomous vehicles are extremely complex systems of software and hardware working together.
Key Functions of Autonomous Driving Systems - How self-driving cars work
At a fundamental level, autonomous vehicles are designed to be capable of 3 things:
- Sensing: Autonomous vehicles must understand their environments – which requires the integration of complex sensors like cameras, radar and LIDAR. A fully autonomous vehicle is expected to incorporate more than 25 image sensors. To accommodate this demand, the market for image sensors is expected to grow by 800% in the next ten years, from 65 million to more than 500 million.
- Thinking: Autonomous vehicles will need to be capable of fusing data from sensors and thinking intelligently through the scenarios within their environments. To realize this capability, engineers must design highly sophisticated integrated circuits and software to enable the vehicle to understand its surroundings, as well as any present variables.
- Acting: Using complex algorithms to control steering, acceleration, and braking, autonomous vehicles must act in a safe, repeatable, and reliable manner. Certifying this behavior will require exhaustive testing with coverage of staggering numbers of use cases. Even with advanced simulation technology, design validation may perhaps be the largest cost component in the development of autonomous vehicles.