The challenge
Fuel consumption rates of trucks transporting freight can swing considerably because of air resistance, or drag. Nearly half of all fuel they consume is used fighting drag, making trucking one of the most energy-intensive modes of freight transportation.
Because of the nature of physics, the level of drag changes constantly because of factors such as the presence or absence of physical objects, for example, buildings and other vehicles, or changing road or weather conditions. These factors can swing fuel consumption dramatically, reducing it by as much as 15% and increasing it by as much as 25%. These real-world complexities; however, cannot be captured with traditional wind tunnel testing.
Researchers began wondering whether vehicles could learn to sense, and adapt to, their aerodynamic environment, much like fish or geese do in nature. But to explore this idea, they would need a different way of capturing the physics. This need gave birth to a new bio-inspired project that could help completely transform freight trucking.
Origins in ocean innovation
Initially funded by the NRC's Ocean program, this project draws on previous work by the Automotive and Surface Transportation Research Centre and is aimed at improving the efficiency of maritime autonomous surface ships (or MASS). Through this work, researchers identified a new pathway for improving MASS efficiency: bioinspiration. This refers to the human development of novel materials, devices, structures and behaviours inspired by properties and behaviours of biological organisms.
The idea behind their innovative approach to improving the efficiency of MASS was simple, but powerful: just as schools of fish conserve energy by swimming in formation and adapting to one another's pressure fields, fleets of ships could perhaps learn to operate as highly efficient swarms. Such swarming behaviour held promise for major energy savings and improved operational performance, in both commercial shipping and military operations.
Researchers at the NRC and Queen's University began by studying how fish sense their surroundings through their lateral line, a sensing system that detects subtle changes in water pressure and flow. The research team outfitted canoes and small boats with arrays of pressure sensors and were able to show that MASS could gain "self-awareness" by feeling hydrodynamic pressure fields.
A change in direction
Dr. David Rival from Queen's University explored the topic in the context of shipping alongside Dr. Taufiq Rahman from the NRC. Together, they realized this concept could be applied beyond the sea. Seeing an immediate opportunity to transform freight trucking, where a reduction in drag translates directly into lower fuel costs and emission levels, they went on to partner with NRC aerodynamicist Dr. Brian McAuliffe.
Bioinspiration in action
These researchers looked to nature as a source of inspiration, specifically at fish and geese, in the context of truck aerodynamics. This "cross-pollination" of ideas led them to an unusual hypothesis:
- Fish adapt to their neighbours using pressure sensing.
- Geese fly in formation to exploit the aerodynamic advantages.
- By analogy, trucks could be made to "sense" airflow and adapt their positioning to minimize drag.
This cross-pollination allowed them to reframe the original project from swarms of ships at sea to swarms of trucks on land.
Proof of concept
Using commercial off-the-shelf sensors, Dr. Rival first demonstrated that pressure sensing could capture aerodynamic states on moving vehicles. Through his subsequent work with Drs. Rahman and McAuliffe, he showed that vehicles could, in principle, "feel" their environment.
Roles and expertise
The research team capitalized on their diverse areas of expertise and the various aspects of the work needed to develop their proof of concept:
- Dr. David Rival applied data-driven machine learning and transition network concepts to model how vehicles switch between aerodynamic states as they seek more configurations.
- Dr. Taufiq Rahman outfitted vehicles with light detection and ranging (LiDAR) technology, video cameras and pressure sensing kits, led the on-road experiments and developed perception and control algorithms for cooperative vehicle positioning within a platoon.
- Dr. Brian McAuliffe, whose earlier research in truck platooning, laid the foundation for understanding and modelling multi-vehicle interaction, used his aerodynamic expertise to analyze air flows in wind tunnels and in real-world scenarios.
Impact
By combining physics-based data with machine learning, the team is pioneering a new approach to self-sensing, adaptive vehicles. The implications are significant. In the trucking industry, this approach could translate into major reductions in emissions and fuel consumptions, while intercity and highway driving could also become more fuel efficient and predictable.
Beyond its applications to trucking and shipping, this research helps reinforce the potential long-term benefits in a number of other areas as well, such as traffic systems designed to improve energy efficiency, all inspired by the schooling behaviour of fish! The innovation offers immediate opportunities for land-based applications and for generating strategic value for Canada's commercial shipping and defence operations.
Next steps
Although the project is still at the exploratory phase, the path forward is clear to them. They will work toward integrating physics-informed, data-driven methods into structured platforms to guide real-world vehicle behaviour.
They will expand on-road testing using outfitted vehicles and refine control strategies for vehicle repositioning. At the same time, they will continue to use wind tunnel data and aerodynamic analyses to inform the models. As proofs of concept mature, the team will be well positioned to scale their methods into robust applications, providing a new generation of adaptive, efficient transport systems.
"This is a great partnership that will provide a lot of very useful information for intercity and highway driving."