Most people don’t realize what goes into hacking an autonomous truck.

But the students in the capstone course taught by Virginia Tech Bradley Department of Electrical and Computer Engineering Professors Almuatazbellah Boker and Ravi Raghunathan do.

Paola Cando Bojorque, Ramida Theeravachirakul, Spurthi Mohan, and Max Aedo Espicto spent their fall and spring semesters dedicated to addressing this issue in their capstone project with Torc Robotics. The goal seemed simple enough: prevent the vehicle from colliding with objects and create pedestrian safety. The journey, though, proved challenging – but nothing they couldn’t overcome.

“At the start of the fall semester, we didn’t have a clear path,” said Cando Bojorque, a Virginia Tech Innovation Campus Master of Engineering in computer engineering student involved in the project. “But we knew we’d be working on a project that involved machine learning and cybersecurity.”

The team was correct. They received their final directive: develop a software to track new threats and vulnerabilities in the autonomous truck’s software and notify Torc staff through a tracking system. The four students first had to consider the systems that allow an autonomous vehicle to function – specifically the laser detection and ranging systems (LiDAR). LiDAR scans around the vehicle and uses the data to inform it of its surroundings: people, other cars, traffic lights, police, road abnormalities (like roadwork), and more.

“That’s where our team comes in,” said Cando Borjoque. “Those systems have to work just right or otherwise a lot can go wrong. These sensors are easy to hack and jam.”

Every autonomous vehicle sends its data through an Internet of Things (IoT), which is a collective network of connected devices and the technology that facilitates communication between devices and the cloud. They also have a security system that utilizes WiFi. Because there is a constant cycle of information sent to the cloud and back to the vehicle, weaknesses are easy for hackers to spot. Normally, cybersecurity departments use information from vulnerability management systems (VMS) to make a list of the technical components of a vehicle, analyze them, identify which software or hardware has a vulnerability, and then patch the issue to ensure the security of the system. Once created, these patches are then uploaded to the National Institute of Standards and Technology (NIST) database.

The student team “dissected what went into a vulnerability,” Aedo Espicto said, and created a faster way to analyze an autonomous vehicle’s security weaknesses. Their VMS automatically connects to the NIST database and another managed by MITRE, enabling the team to discover issues and make patches quickly. Future work includes creating a complimentary system that tracks vulnerabilities and notifies Torc staff of them.

“When you’re in school, you’re turning the concepts you learn into papers,” said Cando Borjoque. “In working directly with a company, you’re developing a project that might be used in the industry. You’re managing stakeholder expectations. You’re working with a team that is pushing toward a mutual goal. At the end of the day, your client needs that project.”

These teams work directly with a representative from the sponsor company. The Torc capstone team was assigned Bill Rushmore, lead cybersecurity engineer. Rushmore served as the project’s subject matter expert, shared what Torc needed from the team, and helped them achieve it.

“Working with the Virginia Tech students was a real pleasure. I was really impressed with how quickly the team came up to speed on the problem that Torc needed to solve,” said Rushmore. “It was especially impressive that they weren’t coming onto the project with cybersecurity in mind, but by the end of the first semester were on their way to becoming experts in cybersecurity vulnerability tracking. Ultimately the team developed something that was beyond just an academic exercise but a solution that we will use.”

Rushmore’s colleagues Wayne Epps and Austin Prosser organized a tour of Torc’s facilities in Blacksburg, VA, where the student team was given a demonstration of how the autonomous vehicles and their sensors operated.

Capstone projects are a component of the Innovation Campus’ project-based learning model. At full build, the Campus hopes the curriculum will be between a third and up to a half of project-based learning. Additional student teams worked on projects with companies like The Boeing Company, the U.S. Marine Corps, and Collins Aerospace. This course is designed for students to build real-world skills within the comfort of a collegiate environment, so when they graduate and begin their first job, they are confident and prepared.

Take it from Cando Borjoque: “Work through the challenges your capstone throws at you and enjoy the process. Sometimes the process is very frustrating, but you’ll learn something about a teammate or about the industry. You’ll learn to ask for help and input. Don't be afraid to go give it your all.”

Share this story