Tracking and Control of Articulated Machines through Remote Sensing
Reference number | |
Coordinator | Örebro universitet - Center for Applied Autonomous Sensor Systems |
Funding from Vinnova | SEK 4 496 653 |
Project duration | August 2017 - November 2021 |
Status | Completed |
Venture | The strategic innovation programme for Swedish mining and metal producing industry - SIP Swedish Mining Innovation |
Purpose and goal
A key enabling technology for remotely controlling mining machines and removing the operator from hazardous environments is the ability to estimate and control all different joints for each arm. In the TracMac project we have designed a completely new type of method to estimate the position of all joints in the arms: using remote depth sensors and AI. By removing all sensors from the arm, we aim to get a more robust hardware that can handle a tough environment. The technology we have developed will be further explored through the use of new sensors specially adapted to a mining environment.
Expected results and effects
TracMac resulted in the development of novel technology for remote sensing and manipulator tracking in harsh mining environments. In the project we have developed software modules for estimating the state of an articulated machine, using remote sensing and machine learning. The resulting technology is expected to increase the robustness of autonomous and teleoperated machines, and drive down maintenance costs.
Planned approach and implementation
The project was planned to run for three years, but was extended due to the covid-19 pandemic. During the first year, we evaluated the basic technology for remote sensing. During the second year, we implemented a prototype tracking system and evaluated it in laboratory conditions. During the third year, the tracking system was improved and augmented with a module for modeling hydraulic hoses. Data collection for evaluation in underground mining was delayed during the pandemic and only completed during the last months of the project.