Expanding Robotic Interaction and Maneuverability Through Holonomic Motion In Remote Robotic Intervention

Authors: Arvin Niro, Makana Ramos, Jeffery Oshiro Faculty Mentor: Dr. Aaron Hanai, Ph.D.

Kapi‘olani Community College, STEM Program

Abstract Robots have become fairly popular when it comes to designing and testing new technologies, and are also prevalent for remote intervention tasks. Sending robots to Mars to explore the surface, or having a bomb squad extract a bomb are some examples of remote robotic intervention tasks. By using Controller Area Networking (CAN) communication, and understanding the concept of holonomic motion, robots can become very versatile and mobile when it comes to intervention tasks.

Introduction Holonomic motion is defined as being able to control all of the degrees of freedom (DOF) that a system has. These degrees can be shown in Figure 1. Figure 2 shows the vector analysis of holonomic motion where it allows a rigid body to move in any desired direction without disrupting its orientation. Because of this special characteristic that holonomic motion allows, maneuvering in tight spaces becomes surprisingly easier. Designs There are two designs that allow for holonomic motion; Omni-wheels and Mecanum wheels. Both wheel designs have individual rollers that are mounted in a certain angle that is opposite in the rolling direction of the wheel. Figure 3 and Figure 4 display what a omni-wheel and mecanum wheel look like.

Figure 3: Omni-wheels Figure 4: Mecanum Wheels

Method This section will cover the following: • CAN Communications • Chassis • Holonomic Wheels CAN Communications For our robot, we are using a kit provided by AndyMark. In this kit we have a Pocket N-Router, 2CAN device, and a Canipede RCM. These devices will allow for other components to easily be intergraded into our robot such as sensors and anthropomorphic robotic arms. Chassis For this project, we are using the AndyMark C-Base Chassis Kit that is used in the FIRST Robotics Competitions. However we designed the chassis in Solidworks before we began to fabricate it as shown in Figure 5. Figure 6 shows the current robot after fabrication has be completed.

Holonomic Wheels In order to achieve holonomic motion, we must first look into the physics of the mecanum wheel. Figures 7 – 9 shows the sum of all forces and moments that act on each mecanum wheel. Adding up all of these vectors, we have a robot that has 3 DOF, which is the total amount of DOF’s that our robot has. (Figure 10)

Figure 7: Translation along Y direction Figure 8: Translation along X direction

Figure 9: Rotation along Z direction Figure 10: Total of 3 DOF

Results Our robot, which uses 10” diameter mecanum wheels from AndyMark, standing at just 9 inches tall, 2 feet wide, and 22 inches long, was able to navigate through tight spaces such as between tables, chairs, doorways, and spaces where it was almost impossible to turn. The wheels provided the operator with lots of space to maneuver and was able to control the robot with ease.

Purpose The goal of this project is to design and build a robot that utilizes CAN communication technology and holonomic motion in order to navigate across campus and retrieve an object while the operator remains in a remote location. To ensure that our robot can maneuver in tight spaces and achieve holonomic motion, we look into the characteristics of mecanum wheels.

Acknowledgements: Big mahalo to our mentor Dr. Aaron Hanai for providing insight and assistance with the project, Dr. Hervé Collin for providing additional assistance, KCC STEM Program for allowing us to use the facilities, and NSF for their monetary support.

Method (Continued) Using what we discovered through vector analysis of the mecanum wheels, we are able to control all movements of the robot, thus can move in any direction by turning each individual wheel in a certain direction as shown in figure 11. In order to move each wheel individually, the operator uses a program called Cross-Link Robot Control System (provided by the CAN system product) that is also connected to a controller. In the program, the operator can assign buttons or motions on the controller in order to move the robot. Once that is done, the operator can connect to the router mounted on top the robot via WIFI or Ethernet cable. The input signals that are given by the controllers and program get sent to the router, to the 2CAN, to the Canipede RCM, and then to the four speed controllers (Jaguars), which are each connected to a motor and wheel, thus turning the wheel in the direction the operator programmed. The onboard webcam sends live video footage back to the operator in order to see where the robot is facing and doing and the operator can make adjustments based on the footage. This process can be shown in figure 12.

Figure 12

Conclusion By looking into holonomic motion, this allows robots to move in any direction which gives the operator full control of where the robot is facing and how it turns at any given time. However there are some disadvantages of holonomic wheels. The biggest problem is that because there are smaller rollers mounted around the wheel, this makes the robot vulnerable to other forces that can cause the robot to move due to the little friction that is given by the wheels that are in contact with the ground. Holonomic wheels also cannot reach higher speeds due to the design of the wheels and the amount of friction generated from each roller at higher speeds. Despite these disadvantages, the advantages outweigh them. For example a company called Airtrax uses mecanum wheels on their forklifts. The forklifts work in the same way as our robot, but instead they can transport bigger goods inside warehouses and not have to worry about reversing and hitting other objects and can transport these goods with ease. When used in intervention tasks, maneuverability is important because the robot is like an extension of a human beings arms and legs. By sending a robot instead of a human to perform a dangerous task, it reduces the risk of a human losing their life.

http://www.worldwideflood.com/ark/anti_broaching/wave_yaw.htm http://www.cs.cmu.edu/~pprk/physics.html

http://www.andymark.com/product-p/am-0903.htm http://www.andymark.com/product-p/am-0583.htm

http://nexusrobot.com/attachment.php?id_attachment=15

Dassault SystemesSolidWorks Corp

Figure 5 Figure 6

Figure 1 Figure 2

Figure 11

Future Research Right now for the robot, we are working on setting up the WIFI communication so that we could use our school’s WIFI and 4G hotspots to allow us to take the robot anywhere. The chassis of the robot will be redesigned and built using FRP unistrut material. Our group is also looking to integrate an anthropomorphic robotic arm, specifically the AX-18A robotic arm made by CrustCrawler Robotics, to allow our robot to interact with the environment. It will also have a webcam mounted to send feedback to the operator. (Figure

13) http://www.crustcrawler.com/products/AX-18F%20Smart%20Robotic%20Arm/

Figure 13