By Emily Jankauski
Let’s face it, car sensors are becoming part of our everyday routine — from alerting us when we’ve veered too closely to another lane to helping us back out safely from our driveways.
For Alireza Talebpour, a Texas A&M assistant professor who will join the ranks of the University of Illinois at Urbana-Champaign’s Department of Civil and Environmental Engineering in January, today’s sensors are a far cry from their true potential.
Imagine sensors alerting you of a pedestrian about to walk out in front of you or helping your vehicle communicate with cyclists.
That’s the future Talebpour sees in his autonomous vehicle sensor research.
“Sensors are the eyes of the vehicle,” he said.
For Talebpour, sensors are the key to unlocking the future of autonomous vehicle exploration.
One of the more exciting problems he’s working on testing is combining cameras with radar sensors, which act as the autonomous vehicle’s eyes most often in low-visibility scenarios such as nighttime.
“Understanding the combinations of those — being able to detect something in the camera and find the same object in the radar data to see how far the object is — that is a real challenge,” Talebpour said.
The assistant professor is also attempting to tackle sensor configuration ― deciphering just how many sensors are necessary to operate the vehicle. Currently, autonomous vehicles have a “ton” of sensors, and their constant use drains the car’s energy.
Let’s say you’re cruising along Interstate 57 with no cars around you. Would it be necessary to have all of the autonomous vehicle’s sensors running?
Probably not.
“You just need one camera, maybe a couple of other ones to tell you where you are, to inspect the lanes, and it’ll get you going,” Talebpour said. “All of the other sensors, if you run them you are just losing energy.”
This means you’ll need to recharge again. And that constant recharging only increases maintenance costs and ultimately shortens the lifespan of the vehicle, according to Talebpour.
Perhaps the most exciting element Talebpour can’t wait to get his hands on is exploring real-world scenarios. He’ll definitely have all kinds of elements to tinker with now.
“This is an excellent opportunity,” he said. “(It) will have new weather conditions, new driving environment(s) and also all of these activities going on in terms of the new test track (the Illinois-Autonomous and Connected Track) that is under development and planning.”
“All of that is quite exciting,” he added, “because I feel like I will be in the right place to develop my research. I’m so excited to begin working with the students, faculty and staff.”
And with all of this excitement, there’s still one question that keeps Talebpour awake at night: “How does my design impact the transportation system overall?”
“If I change the autonomous cars to behave like X and Y, would that impact the congestion? Would that impact the safety around the city? Would that impact the energy consumption?” he asked.
“Once we do that, then the bigger problems will start, like ‘OK, this happened, but now what happens to the congestion? Is it worse? How could it be better?’”
Such existential questions urge Talebpour to think critically in the here and now.
“This is the right time to consider all that,” he added, “during the design period, not after implementation.”
That’s something he hopes to develop at UIUC.
“No one has the flexibility like universities to think freely and try things that are out of the box,” Talebpour said. “Having that ability to think out of the box and implement those ideas can help (create) much safer autonomous driving.”