Oct 30, 2019 Automotive

The Next Phase For Autonomous Cars — Conquering Real World Challenges

By Ty Kim

How is the acceleration of autonomous technologies forcing developers to rethink their intelligent vehicle approach?  This article explores some practical challenges and provides insights on the future of autonomous cars.

The current acceleration of autonomous technologies is leading to unprecedented changes for the automotive industry.  A new reality arises for OEMs — they are racing to keep up with demands for technology in which they traditionally have limited expertise.  What is needed is a new design approach to make autonomous technology safe, scalable, fit timelines and meet budgets.  

In an autonomous car, the vehicle compute workload creates an enormous flood of data that is driven by the proliferation of sensors, ECUs, cameras, radar and smart, connected devices.   By some estimates, each autonomous vehicle will be generating approximately 4,000 GB – or 4 terabytes – of data a day.  Those figures are hard to ignore.

The trend towards workload consolidation platforms using multi-core processors brings entanglement and interference issues that complicate resource provisioning, element separation, certify-ability, security and safety. It's not only hardware resources such as CPU, memory and devices that need to be managed and considered by the run-time environment but also relevant attributes such as determinism, bandwidth, boot-time, certification level, viral license considerations, static vs. adaptable software, dynamic update and other system integration considerations.

The implication is to base such systems on an integration platform that provides for robust partitioning – and to pursue a system architecture that uses the principle of separation of concerns. Designers will look to separate software elements that have differing core attributes.  One solution is to partition mixed criticality using a certified real-time hypervisor that supports fractional core scheduling in addition to robust space partitioning.

The requirements to support complex automotive applications, while allowing flexibility in processing distribution and compute resource allocations, has led to an AUTOSAR adaptive platform with a service-oriented-architecture (SOA).  The services and applications can move in both time and space, meaning it can exist in multiple ECUs in a vehicle as well as outside the vehicle e.g. infrastructure devices of an intelligent traffic system (ITS).

Connected, intelligent vehicles hold great promise and will enable change across a range of industries and define a new landscape.  Let’s start a conversation that explores the catalysts for change, frames the value or economic opportunity, and begins to build use cases that can enable business decision-makers to explore and develop actionable autonomous driving strategies.

To learn more, please explore my ARM TechCON 2019 presentation where I discuss the acceleration of autonomous vehicle technologies that’s forcing developers to rethink their approach and how it is leading to unprecedented changes for the automotive industry.

If you’re interested in learning more about automotive solutions and technologies, then connect with us.

If you found this article useful then please share it with your LinkedIn community.

Previous Evolving Embedded Linux Together
Next The New Space Race Is an Agile Odyssey