Sheridan College and Rogers Communications are collaborating on research into autonomous vehicles.
The two-year partnership will focus on integrating navigation, diagnostics and infotainment systems into autonomous vehicles using wireless 5G technologies.
Sheridan has been working with Rogers for almost three years. But Edward Sykes, the director of Sheridan’s Centre of Mobile Innovation (CMI), sees this research as a new opportunity that has become one of CMI’s main projects since it began this January. “We are thrilled to collaborate with Rogers and looking forward to the future,” he said.
Rogers provides Sheridan’s researchers and students with a 5G platform. The researchers use the platform to create and develop apps and solutions that will use machine learning to study how navigation, diagnostics and infotainment systems can be integrated into vehicles and introduced in a multi-user environment.
The 5G research will tackle autonomous delivery systems, among other autonomous driving challenges. For long-distance routes, for example, navigation and road-mapping technologies can be used to pre-define highways and speed limits, minimizing gaps between the trucks and saving fuel, said Khaled Mahmud, a principal investigator on the Rogers and Sheridan’s project.
Last-mile delivery will still pose challenges, Mahmud said. In some gated areas, for example, completing deliveries without a driver might be impossible, even with small everyday deliveries such as groceries.
Level 4 autonomy
“In this project, we are working on achieving Level 4 of autonomy, where humans don’t have to be present in a vehicle for it to drive. When you are in a car running itself in the road, and there’s nobody’s there, it’s safe,” Mahmud explained.
“But the challenge comes when there are other cars there, and you need that direct car-to-car communication.”
This cellular vehicle-to-everything (C-V2E) technology is critical for vehicles to communicate with each other and with all the devices around them. The newly introduced 5G technology makes it possible for researchers and students to use machine learning techniques combined with traffic modelling and simulations to generate insights into reading, detecting, and processing information, saving it on the Cloud and transmitting it to other devices, Mahmud said.
“When you’re driving, you might not understand from the distance if there’s a human on the road. But with a machine learning algorithm, cars would be able to detect a building, a lamppost, another car or cellphone if there’s a built-in algorithm in it. This technology should be taken advantage of, because it would prevent multiple accidents. But this is a multi-user, multi-device environment, which gets more complex with numerous devices. Its feature, size and capacity get very complex.”
The research is in its initial stage, Sykes said, and it is many years until these technologies will be integrated into our lives.
He added that this research is dedicated to solving questions in just one specific area. “There are so many competing problems that need to be all addressed, many challenges that need solutions to have fully autonomous vehicles. For instance, the battery is a significant problem at this point. Many researchers are exploring different types of chemical compounds to maximize battery performance and prevent overheating. And then, how do you provide a comfortable cabin environment? There are tons of problems and challenges associated with this area. What we are trying to do is identify one slice and chip away on it, and that, in itself, will be a contribution to solving one of these challenges,” Sykes concluded.
Leave a Reply