TY - GEN
T1 - Fast and Accurate Simulation of Mecanum Wheels with Passive Rollers Emulated by Fixed Joints and Anisotropic Friction
AU - Okada, Yoshito
AU - Oguma, Kazuya
AU - Gunji, Kenta
AU - Yokota, Yoshiki
AU - Aryadi, Hanif
AU - Kojima, Shotaro
AU - Bezerra, Ranulfo
AU - Konyo, Masashi
AU - Ohno, Kazunori
AU - Tadokoro, Satoshi
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - Simulating mecanum wheels, ubiquitous in robotics and industry, is a critical tool for advancing research and development on robots equipped with these wheels. This study proposes a fast and accurate simulation model for mecanum wheels, where passive rotating rollers attached to the wheel's outer circumference are represented with fixed joints and anisotropic friction links. The simulation is fast because the constraints on the rotation axis of the roller do not need to be explicitly considered in the physics calculation. The barrel shape of the roller is approximated using multiple spherical links placed along the axis of the roller. As this approach replicates the shape with fewer polygons, simulation accuracy is maintained, and collision detection during the physics calculation can be accelerated. Comparison experiments with conventional models and an actual vehicle confirm that a simulation accuracy close to that of the actual vehicle can be achieved at a calculation speed more than three times those of conventional models. The proposed simulation model may aid in the development of transportation vehicles for factory applications, and it may replace labor-intensive experiments on swarm vehicles to expedite research and development therein.
AB - Simulating mecanum wheels, ubiquitous in robotics and industry, is a critical tool for advancing research and development on robots equipped with these wheels. This study proposes a fast and accurate simulation model for mecanum wheels, where passive rotating rollers attached to the wheel's outer circumference are represented with fixed joints and anisotropic friction links. The simulation is fast because the constraints on the rotation axis of the roller do not need to be explicitly considered in the physics calculation. The barrel shape of the roller is approximated using multiple spherical links placed along the axis of the roller. As this approach replicates the shape with fewer polygons, simulation accuracy is maintained, and collision detection during the physics calculation can be accelerated. Comparison experiments with conventional models and an actual vehicle confirm that a simulation accuracy close to that of the actual vehicle can be achieved at a calculation speed more than three times those of conventional models. The proposed simulation model may aid in the development of transportation vehicles for factory applications, and it may replace labor-intensive experiments on swarm vehicles to expedite research and development therein.
UR - http://www.scopus.com/inward/record.url?scp=85185847441&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85185847441&partnerID=8YFLogxK
U2 - 10.1109/ICAR58858.2023.10406382
DO - 10.1109/ICAR58858.2023.10406382
M3 - Conference contribution
AN - SCOPUS:85185847441
T3 - 2023 21st International Conference on Advanced Robotics, ICAR 2023
SP - 592
EP - 598
BT - 2023 21st International Conference on Advanced Robotics, ICAR 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 21st International Conference on Advanced Robotics, ICAR 2023
Y2 - 5 December 2023 through 8 December 2023
ER -