Collision-Aware Coordination of Dual Drilling Manipulators Using RoboCompiler and Control Barrier Functions
This paper presents a collision-aware coordination framework for dual hydraulic drilling manipulators that integrates RoboCompiler-generated mechanism models with centralized Control Barrier Function optimization to achieve safe, efficient concurrent drilling, demonstrating a 42.5% reduction in mission time compared to sequential operation while maintaining strict clearance and contact constraints.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Deep underground, where the air is thin and the rock is hard, machines must work with a precision that human hands can rarely match. In the tunnels of mines and the foundations of cities, massive hydraulic arms drill holes to stabilize the earth or prepare it for blasting. These machines are powerful, but they are also slow to react. When a hydraulic cylinder receives a signal to stop, the fluid inside does not halt instantly; the metal continues to move, carrying momentum that can push the tool past its intended mark. In a confined space where two such arms operate side by side, this delay creates a dangerous gap between what the computer orders and what the machine actually does. If the arms move too fast or too close, they can crash into each other or into the tunnel walls, causing costly damage or halting work entirely. The challenge for engineers is not just to tell the arms where to go, but to predict exactly how far they will travel before they can truly stop, ensuring they never violate the invisible safety bubble around them.
A team of researchers at Tampere University and Sandvik Mining and Construction has developed a new way to manage this delicate balance. They created a system that allows two drilling arms to work at the same time in the same tight space without colliding. Instead of treating the two arms as separate machines that happen to be near each other, the system views them as a single, coordinated unit. It uses a mathematical safety filter that constantly checks the distance between every part of both machines, the tunnel walls, and the rock face. Crucially, this filter does not just look at where the arms are right now; it calculates where they will be in the next few moments, accounting for the lag in the hydraulic systems and the time it takes for the machine to brake. By reserving a specific amount of space for this stopping distance, the system ensures that even if the machine is moving at full speed, it can always come to a safe halt before a collision occurs.
The researchers tested this approach using a detailed computer simulation of a drilling machine. They set up a scenario where the two arms had to drill twelve holes in a specific pattern on a rock wall. In this test, the arms worked together, moving simultaneously to their targets. The system successfully guided the arms through the entire mission, drilling a total of 48.035 meters of holes. The arms never touched each other or the tunnel walls, maintaining a safe distance even when they had to move past one another. In fact, the simultaneous operation was 42.5 percent faster than if the arms had taken turns drilling one after the other. The system also handled unexpected hurdles; when one arm got stuck or needed to change its path, the other arm would automatically slow down or pause to let it pass, then resume its own work once the path was clear.
To prove the system was robust, the team ran a much larger test involving one hundred separate docking operations. In these trials, the arms had to move to a specific spot and press against the rock with a precise amount of force, simulating the moment a drill bit first touches the surface. The system succeeded in 294 out of 300 attempts. In the few cases where it failed, the issue was not a collision, but rather a difficulty in finding a path to the target due to the complex geometry of the tunnel. Throughout the successful trials, the arms approached the rock at speeds well below 20 millimeters per second, ensuring a gentle and controlled contact. The researchers also compared their new system against older methods that did not account for the hydraulic delays or the need for a coordinated stop. The older methods were either slower or, in some cases, allowed the arms to get dangerously close to each other, sometimes even touching when the computer thought they were safe. The new system, by contrast, kept a consistent safety margin, proving that understanding the physics of the machine's movement is just as important as knowing where it needs to go.
The success of this work relies on a clever integration of two different technologies. One part of the system is a tool that builds a detailed map of the machine's physical structure, understanding how its joints and cylinders connect. The other part is the safety filter that uses this map to calculate the safe limits of movement in real time. By combining these, the researchers created a controller that can make split-second decisions about speed and direction. It decides how fast each arm should move and when to slow down, all while ensuring that the total amount of hydraulic fluid being used does not exceed what the machine's pump can provide. This coordination allows the machine to be both efficient and safe, moving quickly when the path is clear but instantly adapting when the situation becomes tight.
The results suggest that this approach could significantly improve the efficiency of underground construction and mining. By allowing two large, heavy machines to work in harmony rather than taking turns, the time spent on drilling tasks could be cut by nearly half. More importantly, the system provides a layer of safety that is difficult for human operators to maintain consistently, especially in the dusty, noisy, and visually obscured environment of a tunnel. The researchers confirmed their findings through rigorous testing, including cross-checking their simulation models with a different physics engine to ensure the results were not merely an artifact of the software. While the work was conducted in a virtual environment, the detailed modeling of the hydraulic systems and the physical laws governing the machine's movement suggests that these results would hold true in the real world. The study demonstrates that with the right mathematical framework, complex machines can be made to work together safely, turning a potentially chaotic environment into a space where precision and power coexist.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.