How technologies from smartphones help realize the dream of autonomous driving?
Level 2, level 2+, level 2 ++… where are we heading? Will we ever see level 3? It seems that trying to segment the automobile industry by autonomy level is not only a difficult exercise, but one with questionable relevancy. After all, isn’t this just another example of a logical, temporal evolution of technological improvement like we have seen with every other new technology – and thus not ideal for a “level-by-level”, incremental approach? Wouldn’t an ideal segmentation just consider everyone’s car as either “standard” or “ADAS”, and mobility as a service from another based on robotic cars?
To answer to these questions, Yole Développement (Yole) has focused on sensor challenges first, and then on the associated computing required for new features/improvements. Lastly, we have tried to organize these topics in cases. During this webcast, Yohann Tschudi, Market and Technology Analyst in Computing and Software at Yole, will explain the evolution of computing and how it has been notably impacted by the arrival of AI (or more precisely, deep learning), and what we expect to happen. For example, AI’s impact is characterized by the entry of technologies from the smartphone industry, such as the neural engine and the accelerator, which indicates a desire to increase performance without increasing consumption. Does this mean that we will need a new type of computing, and new architectures?
Finally, where is the finish line for this race? What technologies and players are involved? And perhaps most importantly, who can win?
For Stéphane Cordova, Vice President, Embedded Technology Business Unit at Kalray, the Automotive industry is currently facing two major challenges: a need for performance and a need to consolidate the electronic functions in the car:
- Growing need of performance for autonomy for car perception and path planning for example
- A way to discontinue the past 20-year way to add functions: add yet another ECU (Electronic Component Unit) for yet-another function
He will also focus on performance and aggregation of heterogenous functions, ensuring mandatory and high levels of security and safety are the keys for upcoming autonomous vehicle production.
Do not miss this webcast to get a deep anlysis on the AI computing for automotive!
Related reports:
Artificial Intelligence Computing for Automotive 2019
Artificial Intelligence for automotive: why you should care.
2020 update to be released in May 2020!
