Concept Reply, a provider of IoT solutions, has partnered with Intel to launch an AI-driven innovation aimed at enhancing the safety and efficiency of autonomous vehicles. Drawing on their combined experiences with AI and AVs, the solution enables precise traffic light detection, allowing it to bolster road safety in urban environments.
At the core of this solution is CARLA, a platform designed to simulate complex car sensor functionalities – including GPS, LiDAR, and accelerometers. Here, it offers simulations for various driving scenarios – such as night time driving and diverse weather conditions – in an approach that allows the companies to elevate the precision of the solution and fine-tune its algorithms.
Ensuring the reliability and precision of CARLA’s robust capabilities demands extensive machine learning training that, itself, requires advanced GPUs. However, in light of today’s GPU shortage, Concept Reply partnered with Intel to seamlessly integrate CARLA with cloud instances optimized by the computing company. This combined platform can scale dynamically, accommodating varying numbers of vehicles during peak hours while efficiently conserving resources during quieter times – running with sustainable computing in mind.
By pioneering this AI-driven solution with Intel, Concept Reply aims to accelerate the advancement of AV safety. Beyond its application in self-driving cars, the joint system can also benefit individuals with color blindness or color vision deficiencies (CVD), contributing to safer roads for everyone.
