Autonomous Solutions
Autonomous trucks are often assumed to stop as soon as the weather deteriorates. In practice, snow, rain, ice, and dust do not automatically prevent autonomous operation. The important question is whether the complete transport solution has been designed and validated for the conditions it will encounter.
- Challenging weather does not automatically stop autonomous trucks.
- By designing and testing the complete transport solution for the conditions customers face, we can keep trucks operating safely through snow, rain, ice and dust.
- If conditions are outside of what we expected, the truck slows down or stops in a controlled way.
- The goal is not to drive through every possible weather event, but to keep production running safely and reliably in demanding conditions.
At a mine or quarry in northern Scandinavia, sub-zero temperatures, snowfall, and icy roads are part of normal operations. They may look challenging or even extreme from the outside, but the transport system is prepared for them. Keeping production moving starts with understanding that environment, defining clear operating limits, and combining vehicle capabilities with site processes and human oversight.
Designing for the conditions our customers face
Every autonomous transport solution operates within an operational design domain, or ODD. This defines the conditions in which the system is designed to work, including factors such as temperature, visibility, and road friction.
Weather becomes “extreme” for the solution when it moves outside that domain. Snowfall does not prevent autonomous operations as the system has been designed to handle it, but a blizzard that reduces visibility below an established limit may do so. Just as it would for a driver of a conventional truck. The same principle applies to friction and temperature.
The ODD is therefore determined by the customer’s real operating conditions. A solution intended for northern Sweden or Norway must be prepared for sustained cold, snow, and ice. And a solution designed for warmer climates must be equally prepared for dust and sand. Accounting for those conditions from the outset creates a broad operating range rather than a system that works only in ideal weather. In Sweden we have an expression, there is no such thing has bad weather, only bad clothing. Essentially, as long as you are prepared, weather isn’t a problem.
A layered response
An autonomous truck uses several sources of information to understand its position and surroundings. LiDAR supports localization and perception, GNSS provides satellite positioning, and radar contributes to obstacle detection. Weather resilience does not depend on one sensor performing perfectly in every condition. It comes from combining technologies, monitoring which capabilities remain available, and adapting the trucks behavior when conditions reduce their performance.
Some challenging conditions are not obvious. On a dry, sunny day, an oncoming vehicle can create a dense cloud of dust. This can affect LiDAR performance and generate returns that must be distinguished from solid objects. The system compares data from LiDAR and radar and adjusts how much weight it gives each. It also analyzes how strongly the points detected by LiDAR reflect the LiDAR’s laser beam, helping the truck distinguish a cloud of dust from a solid obstacle.
Even within the ODD, weather or dust can temporarily reduce the performance of an individual sensor without making the entire system unavailable. The truck continuously monitors whether it still has the capabilities needed to operate safely. If LiDAR-based localization reduces while GNSS positioning remains reliable, the truck can still determine its position, but it has lost one of the two sources it normally uses to navigate. It can therefore reduce its speed. If the remaining information is insufficient for safe operation, the truck stops. What matters is whether the system retains enough reliable information to continue safely.
Weather resilience goes beyond the sensor suite
Challenging weather conditions affect more than visibility. Frost can form on sensors, components can take longer to reach operating temperature, and road friction can change quickly.
The base truck and autonomous driving kit must therefore be considered together. Volvo trucks already operate across a wide range of climates around the world, giving us established knowledge of how the vehicle and its components perform in demanding conditions. This knowledge allows us to specify the vehicle for an ODD that covers a range of customer operating environments. The core specification can therefore serve multiple sites, while certain choices, such as tires, can be tailored to specific local conditions. We then validate the autonomous transport system, including hardware, sensors, and startup procedures under the same conditions so that the complete solution works as intended.
Driving behavior can also be adapted. On an incline with reduced friction, the gearing and control strategy must support reliable progress. On gravel roads, varying the truck’s lateral position instead of following the same track on every cycle distributes wear, reduces rutting, and extends intervals between maintenance activities.
The site is part of the autonomous solution
Continuous production does not depend on the truck alone. Operators, road-maintenance teams, and site-control systems all contribute to keeping the operation running.
Our experience with current customers illustrates the importance of this cooperation. Snowplowing, sanding, salting, and grading are coordinated with the transport operation. Clear instructions and close communication help keep the autonomous operating corridor available.
Human oversight is part of this layered approach. When the truck determined that it cannot continue safely, it stops automatically. The site-control system gives the operator visibility into the event so that the appropriate response can be coordinated, such as clearing an obstacle, inspecting a sensor, or arranging road maintenance.
This combination of automated behavior and operational support provides resilience. Autonomy does not remove people from every decision. It uses their judgement where it adds the most value, while removing drivers from repetitive and potentially hazardous transport tasks.
Combining simulation with real-world testing
Challenging weather conditions cannot be reproduced on demand, so development combines simulation, physical testing, and customer-site data.
Many weather-related scenarios can be assessed in simulation, including conditions involving rain, snow, fog, or reduced visibility. Recorded data from test tracks and customer operations can be replayed to evaluate how the virtual driver responds. Parameters such as visibility and obstacle placement can then be varied to create additional scenarios. This allows us to explore more variations than would be practical or safe to reproduced with a physical vehicle.
However, physical testing remains essential on real trucks with the autonomous driving system, to see how the full system performs. Test-track results show whether simulated behavior corresponds with reality and help improve the realism of simulation tests.
A rigorous solution
Every deployment adds to our understanding. Experience with frost has improved sensor heating and startup routines. Road deterioration has led to changes in driving paths. Dust clouds during summer have driven improvements in obstacle detection, while every new mining environment comes with new challenges on the whole system.
The solution is tailored to each customer’s operating conditions, but it’s not rebuilt from the beginning from every deployment. Data, technical improvements, and operational lessons from one site strengthen the underlying product and can be applied at future sites. As our experience grows across climates and applications, the solution becomes easier to adapt while still being validated for each customer’s specific environment.
There will always be conditions in which stopping is the safest response. A successful autonomous transport solution does not attempt to drive through every possible weather event. It operates reliably across conditions the customer regularly faces, recognizes when essential capabilities are degraded, and responds in a controlled way.
Conditions that appear extreme from the outside can be part of normal operations when they have been anticipated, designed for, and validated. Through a clearly defined ODD, rigorous testing and close cooperation with the customer, we can keep our autonomous solution working safely through demanding conditions and provide a structured route back into production when operating limits are reached.
