An unusual teaching tool has appeared on the campus of Northeastern University in Seattle (USA): an electric Ford Mustang Mach-E covered in sensors. At the wheel is Xian Li, a professor of electrical and computer engineering who teaches a course on robotic sensing and navigation. He admits that he drives his own car aggressively, but this Mustang is driven cautiously: it has expensive equipment on board.
The car is equipped with four cameras, a lidar (a scanner that builds a three-dimensional map of space using laser pulses), a radar for measuring speed and distance to objects, a GPS navigator, an inertial measurement unit (IMU, a device that determines acceleration, altitude and other motion parameters) and eight terabytes of built-in memory. All of this is mounted on the roof, in the cabin and in the trunk. The car weighs about 2,270 kilograms.
According to Li, the Mustang collects images and navigation data at a rate of 10 gigabytes per minute. That is enough for students interested in robotics, computer vision and autonomous driving to test their own navigation algorithms. "Not all students have access to equipment like this," says master's student Xinlei Li (who shares the professor's surname) and who helped install the sensors in spring 2026. "I always wondered whether I would ever get to drive something like this, so for me this is cool."
Autonomous driving technologies allow the car to analyze its surroundings and make decisions: stay in its lane, recognize road signs, and notice a pedestrian in time. Different sensors collect data on traffic, obstacles and the movement of neighboring cars. The difficulty is understanding the strengths and weaknesses of each type of sensor and learning to combine them for the most accurate possible picture of the world, Xinlei Li explains.
Professor Li notes that even advanced Tesla systems with the Full Self-Driving (Supervised) feature, according to the classification of SAE International (the international association of engineers), correspond only to the second level of autonomy out of six, from zero to five. At the fifth level, he says, a passenger could sleep behind the wheel. But such cars do not yet exist, and in the next decade, Li believes, they will not appear. In addition, before fully driverless cars reach the market, the corresponding legislation will have to be created.
The Ford Mustang Mach-E itself does not have full autopilot: it has the BlueCruise system, which belongs to SAE level two and helps keep the car in its lane and manage cruise control while the driver watches the road. However, Li and his students are trying to "hack" the car and introduce more advanced features to bring it closer to level three. They are tuning algorithms in the steering and motor control systems, hoping one day to make the Mustang fully driverless.
The skills gained from collecting sensor data will also be useful outside transportation. Xinlei Li has already used navigation data for a technology that helps visually impaired people navigate unfamiliar places. Another student, Vittoria Biton, studied how images of the environment can help search-and-rescue operations in poor visibility. She proposed adding a neural processor to the car to recognize complex images and creating a perception system that builds a three-dimensional model of the surroundings in real time.
Another student, Elvis Madiba, worked as a programmer in Nairobi (Kenya) before his studies. He says the atmosphere in Li's course, where mistakes are considered part of learning, inspired him to think about introducing a similar model in his home country. "You don't necessarily always have to be 100% right, but if you put in the effort, you can always succeed," he believes.