Calibration of Mobile Robots Using ATOM

Area of Science:

• Robotics
• Sensor Fusion
• Calibration Techniques

Background:

• Accurate calibration is crucial for mobile manipulators, involving sensor-to-motion and localization transformations.
• Existing methods lack flexibility for systems with multiple sensor types.

Purpose of the Study:

• To develop a flexible calibration approach for multi-sensor, multi-modal mobile manipulators.
• To simultaneously estimate sensor poses and calibration pattern positions, even with imprecise localization.

Main Methods:

• Integration of a novel calibration approach into the Atomic Transformations Optimization Method (ATOM).
• Formulation of calibration as an extended optimization problem to estimate multiple transformations concurrently.
• Validation using simulations and real-world case studies.

Main Results:

• Demonstrated improved calibration accuracy for mobile manipulators.
• Showcased effectiveness across diverse sensor modalities.
• Validated the robustness of the ATOM framework.

Conclusions:

• The proposed ATOM-integrated calibration method offers enhanced precision and flexibility.
• This approach is effective for complex mobile robotic systems with varied sensors.
• It addresses limitations in current mobile manipulator calibration techniques.