Session: 08-03-01: Design and Control of Robots, Mechanisms and Structures I

Paper Number: 166132

Design and Development of a Locomotion Compensator for Position Tracking of a Large Insect Using Model Predictive Control 

A locomotion compensator for an insect is an automated robotic system where a sphere rotates in the desired direction to maintain a target insect’s position within the field of view (FOV) of the camera while it walks on the sphere. This device is utilized to observe and track insect behavior ranging from a shorter duration to a longer-term study. Our previous design utilized an omnidirectional sphere-based system that addressed position and orientation compensation but still struggled with the precise tracking of larger insects with various movement patterns. Also, the control of orientation produces undesirable vibrations and slip which results in unstable tracking performance in some cases. The limited size of the sphere restricts the dimensions of insects that can be tracked using the device.  

This paper introduces a modified Locomotion Compensator design with independent control for the x- and y-axis but without orientation correction capabilities. Two motors, a trackball sensor, and a near-infrared (NIR) camera are used for the new design. The motors independently compensate for the insect’s motion in the x and y direction, respectively. The trackball sensor detects and measures the movement of the rotating sphere to determine displacement in the x and y directions.  The camera is fixed on top of the system to observe the insect as it moves on the sphere. The new design also offers the opportunity to use a larger sphere, which enables tracking of larger insects. The redesigned system was used to track large insects such as the pale green assassin bug (Zelus luridus) and boxelder bug (Boisea trivittata), which display more erratic movements compared to smaller insects like ants or fruit flies. Tracking larger insects requires a motion compensation approach that focuses on accurate translation while preventing instability due to rotational corrections.  

We achieved effective control through a Model Predictive Control (MPC) with predicted trajectories of a walking insect using moving averages from previous movement cycles to forecast future positions. The precise control helped to keep the insect within the camera’s FOV while the camera aided in high-resolution imaging of the tracking.  The experimental results reveal that the developed system achieves superior tracking precision by keeping minimal distance error for majority of the tracking duration, which marks a significant improvement compared to conventional controls and control devices with omnidirectional control. Our research shows that eliminating orientation compensation decreases system vibrations along with slippage which leads to better tracking consistency. 

The newly developed system uses a larger sphere along with independent position control mechanisms to handle larger insects while maintaining high-resolution imaging stability. The new development creates a compelling foundation for detailed analysis of insect movement and behavior while showing promise for applications in entomology and neurobiology as well as biomechanics research.  

Presenting Author: Kaushik Rahman Kennesaw State University

Presenting Author Biography: Kaushik Rahman is a Ph.D student at Kennesaw State University.

Authors:

Kaushik Rahman Kennesaw State University
Jonathan Jackson Kennesaw State University
Robert Henderson Kennesaw State University
Josephine Phillips Kennesaw State University
Dal Hyung Kim Kennesaw State University