Picture this: A platoon of logging trucks makes its way through remote forest roads, each responding independently to the environment. A drone buzzes overhead, feeding information to the vehicles. Here's the thing: only the lead truck is driven by a human. The others, equipped with sensors and controllers, drive themselves.
It sounds like a scene from a sci-fi movie, but for a researcher and her team at UNB, this high-tech system could soon find its way to New Brunswick's back roads.
Sensor-driven autonomous vehicles adjust in real time to terrain, wildlife and weather, with the potential to transform forestry, agriculture and oil logistics.
When Dr. Yukun Lu came to UNB, she brought with her a background in autonomous vehicle control and a clear goal: to apply her expertise to a challenge that matters in New Brunswick.
That challenge is the shortage of skilled truck drivers in the province's forestry sector, and the solution she's researching is a human-led autonomous truck platooning system designed to make logging transport safer, more efficient and sustainable.
A platooning system in transportation is a technology that enables a convoy of vehicles to travel closely together in a co-ordinated manner using automated driving systems.
Lu is an assistant professor at the UNB faculty of engineering and the director of the Intelligent Mobility and Robotics Lab (IMRL). Her research focuses on a system in which a human-driven truck leads a convoy of autonomous log-carrying follower vehicles. These followers mimic the lead truck's movements but also make independent decisions based on road conditions, vehicle load and environmental hazards.
"The human driver is still the boss and retains full control of the system," Lu said, adding that fully autonomous vehicles are not yet advanced enough to deal with the challenges that would arise on forest roads, which are often remote, rough and hazardous, especially in winter and spring.
Each autonomous follower vehicle is equipped with sensors, such as LiDAR (Light Detection and Ranging) and cameras, to detect potential obstacles on the road. If a hazard is detected, the follower vehicles can stop independently, even if the lead truck continues. Once the hazard clears, they resume and catch up.
Lu's lab is developing adaptive algorithms that adjust the following distance between vehicles based on terrain, weather and load. For example, on icy roads or steep hills, the system automatically increases the distance to reduce risk.
The vehicles can switch between different dynamic control systems depending on whether they are empty or fully loaded, using technologies such as electronic stability control, torque vectoring and differential braking.
Communication between vehicles is key. GPS is used as a secondary tool to support perception and navigation systems, even when visibility is poor.
In areas where GPS or data reception is unreliable, the system uses direct vehicle-to-vehicle communication and advanced perception (the cameras and LiDAR) to keep all vehicles co-ordinated and moving safely. More work is planned to help set up stable communications in challenging environments.
Lu's research also addresses a less obvious challenge: human behaviour. Her team has found that copying the lead driver's action in all circumstances can be risky if the driver is fatigued, stressed or distracted. The system must monitor the driver's behaviour and adjust accordingly. If the driver's conduct is deemed unsafe, the vehicles following may increase their distance or reduce speed.
"Directly copying the human driver's maneuvers could actually introduce some risk," Lu said. "Our system needs to be smart enough to adapt to human uncertainties."
The lab is testing the system using scaled-down electric vehicles equipped with sensors and onboard computers. These tabletop-sized models allow the lab to run experiments safely in controlled environments.
The next step is to integrate drone technology for aerial monitoring, which would help in situations where the terrain blocks the line of sight between vehicles. Drones could also support mapping and co-ordination tasks, making the system more robust.
Lu sees the potential for this technology to transform forestry logistics.
A single driver could lead multiple trucks, each carrying a full load, reducing the need for additional drivers and addressing labour shortages. The system could also be adapted for other industrial sectors, such as agriculture or oil transport.
Lu is hoping to find partnerships with forestry companies, logistics operators and government agencies to move the research closer to real-world applications. Industry feedback and policy support are essential to bridge the gap between lab testing and commercial deployment.
"We're trying to make people trust human-in-the-loop autonomous driving technology and to see how far this approach can take us toward full autonomy."
