Autonomous Solutions
When autonomous vehicles are discussed, attention often goes to sensors, software, and decision-making algorithms. Those elements are essential, but they are only part of the equation. The performance of an autonomous solution also depends on the base vehicle and how it integrates with the self-driving technology.
That is why autonomy cannot be treated as a software layer added on top of any vehicle. It has to be built on a deep understanding of vehicle dynamics. With the right vehicle and a deep understanding of its dynamics, autonomous systems can make earlier, more precise, and more consistent use of the truck’s capabilities.
Cecilia Ekström is Head of Product Autonomy Enabled Vehicles and Machines at Volvo Autonomous Solutions.
Kristoffer Tagesson is Global Technology Manager Vehicle Motion Management at Volvo Technology.
What is vehicle dynamics?
Vehicle dynamics is the science of how a vehicle moves and responds to driver inputs and to its environment. In practice, it is about how a vehicle accelerates, brakes, steers, and maintains stability.
For heavy trucks, this is especially complex. With up to 60 tonnes of material loaded on a single truck, questions about suspension, braking, and steering become increasingly challenging. Heavy trucks are also fundamentally different from passenger cars, with different chassis designs, axle configurations and load distributions.
For autonomous operation, that behavior must not only be understood in detail. It also has to be predictable enough to model, so the system can plan ahead and control the vehicle with confidence.
Combining that understanding with self-driving technology, the virtual driver can make better use of the truck’s capabilities than even a highly skilled driver could achieve consistently in demanding conditions. In practice, that creates the potential to deliver the equivalent of many highly capable drivers across the fleet.
But it also raises the stakes. If the vehicle is not correctly specified or the strategy is not properly tuned, those shortcomings can scale across the operation just as quickly. This makes vehicle dynamics extremely important.
Matching the vehicle to the mission
This is where the next part of the challenge begins. In autonomous operation, every maneuver still depends on the truck’s ability to execute it under real operating conditions. To do that reliably and efficiently, the vehicle itself needs to be matched to the mission.
That starts with understanding the mission itself: the route, the surface conditions, the payload, the required productivity, and the steepest grades the truck must handle. In mining and quarry operations, gradeability is often a critical factor. It puts demands on influences axle configuration, traction, powertrain choices, and how much power the truck can deliver uphill under load.
There are also challenges when the truck is moving downhill. The truck does not only have to slow down but also manage braking energy without overheating the service brakes. That is why heavy trucks often depend on auxiliary braking systems such as engine braking and retarders.
Conditions under the wheels can also vary significantly. On gravel, mud, snow, or loose rock, traction will vary. In those situations, capability depends not only on torque, but also on how the truck uses functions such as differential locks and axle load distribution to make the most of available grip.
Gear strategy is another critical part of the setup, and we do not simply reuse a standard gearbox strategy in an autonomous application. Volvo has for instance specifically adapted the autonomous Volvo FH’s automated gear-shifting logic to work with autonomous driving, using preview data and intended driving strategy to optimize how the truck prepares for what comes next. That makes it possible to prepare the right gear earlier, maintain momentum uphill, and control speed more effectively downhill.
Getting this right matters even more in autonomous operation, where a truck that becomes struck can be more complex to recover than in a conventional site operation. That makes correct vehicle specification and strategy essential from the start.
How the virtual driver and the vehicle work together
Once the vehicle is configured for the task, the next step is to ensure that it performs as intended throughout the whole operation, and that depends on close coordination between the virtual driver and the base vehicle.
The virtual driver handles perception, intent, and planning. It uses information about the route, slope, intended speed, and the immediate path ahead to decide what the truck should do. The base vehicle then determines how to execute that request based on its configuration and conditions of the road through braking, gear selection, power management, and other motion-control functions.
This is a continuous exchange. The virtual driver provides preview and motion request, while the base vehicle feeds back information about current status, available performance, and any limitation or degraded conditions. That allows the system to adapt in real time as conditions change across the shift.
If we look at gear shifting again, the system is able to anticipate what lies ahead and knows the intended driving strategy and can therefore prepare the right gear before the truck reaches a steep incline. Braking is another. The virtual driver may request a certain level of deceleration, but the base vehicle determines how to deliver it, balancing service brakes, engine braking, and retarders according to the truck’s configuration and the situation at hand.
Just as important, the system must recognize when conditions move beyond the intended operating limits. Autonomous performance depends on using the truck’s capabilities well, but also on recognizing physical limits in traction, braking, and stability.
The result is not just automation, but tighter integration between planning and execution. That is especially valuable in mining and quarry operations, where conditions vary and precision has a direct effect on fuel consumption, tire and brake wear, cycle times, and the ability to maintain momentum over the full transport route.
A complete system
The real challenge for autonomy in mines and quarries is not just making the virtual driver more capable. It is creating a system in which the virtual driver and the base vehicle are designed to work together.
That requires deep knowledge of vehicle dynamics and the ability to learn quickly in the field. When teams can test, tune, and refine the interaction between the vehicle and the autonomous system based on real site conditions, they can improve everything from braking and gear selection to traction strategy and stability control. This approach has been central to our work at our customer sites in Norway and Sweden. And it becomes even more important in critical moments such as raising the bucket, where rollover risk, changing load distribution, and ground conditions demand tight integration between lifting functions, vehicle control, and perception. In these situations, autonomy depends on understanding not just what the truck should do, but what it can safely do in that moment.
That is why the real opportunity in autonomy lies not only in self-driving software, but in building a complete, well-integrated system around the vehicle. That is also how you avoid scaling one bad decision across an entire fleet and instead deliver the equivalent of many highly skilled drivers, operating with greater consistency.
