Key takeaways:
- Complex landscape and market trends analysed
- Key sensor technologies: cameras, lidars and radars
- Market players
- 20-year market forecast
The arrival of autonomous driving will revolutionise the way people travel. Autonomous cars could liberate people from the driving tasks and potentially enhance road safety and efficiency. Autonomous driving (AD) will also improve travel convenience for those who are unable to drive. Currently the autonomous driving system cost is still very high but with the growing maturity of key technologies such as lidars, radars, cameras, artificial intelligence (AI) software and specialized computers, it is expected the cost of autonomous cars will drop significantly in the coming decade.
Mobility services enabled by autonomous driving technology, which allows fleet operators to eliminate the biggest operation cost - the human driver - will offer a cheaper alternative to purchasing and owning a private car. In the next two decades, we expect mobility-as-a-service (MaaS) will grow rapidly to meet the increasing travel demand and in the meanwhile gradually replace private driving. In this class, we will address the peak-car scenario in the context of autonomous MaaS adoption and how it is going to impact on the automotive industry. Key market players in autonomous driving will be analysed with their latest developments introduced.
Perception technology is critical for ADAS and high-level autonomy. The required sensor suite can be diverse, including camera vision, night vision, lidar, and radar. In this class, we cover the following:
Camera: we briefly touch upon various emerging camera technologies that are expected to support ADAS and high-level autonomy functions. These include high-res global shutter camera technologies, and NIR and SWIR sensors using silicon or silicon-hybrid technologies
Lidars: we cover lidar technologies in depth. Here, we consider all the technology choices that must be made in making a lidar and explore their consequences. We cover various laser source and photodetector technologies. We discuss, in length, the various existing and emerging beam steering technologies that exist today. These include rotating mechanical, MEMS, optical phase array, flash, liquid crystal, and many others. We discuss the operation principle, the merits, the design and production challenges, the readiness level, and the current and potential performance level of each beam steering technology. We also discuss market trends, looking at investment trends segmented by technology options. Finally, we provide our technology roadmap, offering our view and insight into how the technology landscape will evolve in the next decade.
Radars: radar technology is already used in automotive to enable various ADAS function. In this talk, we consider multiple technology trends and their consequences. In particular, we examine the shift towards higher frequencies, the shift towards small-node CMOS or SOI technology, high on-chip function integration, trends in packaging technologies, substrate material requirements, and the rise of large antenna arrays. We consider how these emerging trends will improve the angular, range, and velocity performance of radars and densify their point clouds, enabling the training of deep learning algorithms. These trends will lead to 4D imaging radar which can be considered an alternative to lidars.