A hybrid wheel-leg mobile robot.

Doroftei, Ioan ; Marta, Constantin ; Hamat, Codruta 等

1. INTRODUCTION

One of the merits of walking vehicles is their superior terrain adaptability and also stepping over obstacles without applying any forces on them. The use of legs is convenient for locomotion on soft ground or in unstructured environments, where the performance of wheels and tracks are considerable reduced. On the other hand, they are moving with smaller speed comparing to traditional wheeled platforms. Wheeled robots also provide sufficient robustness, mechanical simplicity and energetic performance. They are fast, powerful in terms of load to weight ratio, stable, and easy to control. One way of improving the drawback of walking robots may be to develop hybrid, in order to exploit merits of both walking and wheeled systems.

From an historical point of view, a vehicle using both legs and wheels was not new, because the wheel barrow, bullock cart, etc. have been used since the early days of civilization. But, legged animals were an integral part of these systems and the design of fully artificial hybrid systems started much later Kar, 2003). What can be observed from these systems is that the work traction is done by the legs that allow a better adhesion with the soil, while most of the weight is carried by the wheels. Based on this concept, there are several researches on wheel-leg hybrid robots. In Japan has been developed a hybrid locomotion vehicle for nuclear power plants. It was a vehicle with five locomotion devices, each device consisting of a wheel and a leg., for nuclear power plants. A hybrid robot called SAPPHYR, with two free rear wheels and two traction legs in the front, was designed in France (Guihard et al., 1995). (Muscato & Nunnari, 1999) developed in Italy a robot with two pneumatically actuated front legs, each with three degrees of freedom, and two rear wheels, actuated independently. At University of Pennsylvania has been developed a hybrid wheelchair with two d.o.f. planar legs, for use on uneven terrain, and four wheels. One of the original ideas has been proposed by Hirose, in Japan. He developed a hybrid mobile robot called Roller-Walker with a special foot mechanism in each leg, to change between foot sole for walking mode and passive wheel for skating mode. Researchers at Tohoku University used the idea of combination of four legs (two in the front and two rear legs) and two wheels in the middle, to develop a leg-wheel type robot for the exploration. Another example of the hybrid system, called Walk'nRoll, has been developed by Adachi (Adachi et al., 1999). The robot has two legs with 3 d.o.f. in the front and and two rear wheels. The system usually moves on the surface with wheels like an usual wheeled vehicle. Entering to rough surface area, it switches the moving mode from wheel to leg mode. One of the problems of such systems is how to switch the mode according to the change of surface conditions. To do it, the robot is using ultrasonic sensors. A leg-wheel hybrid robot, with two active 2 d.o.f. legs in the front and two passive rear wheels, has been developed in Thailand (Suwannasit & Laksanacharoen, 2004). In this paper the design of a small wheel-legged robot will be presented. This report is the result of a research conducted at the Robotics Laboratory, Mechanical Engineering Faculty, "Gh. Asachi" Technical University of Iasi, Romania.

2. ROBOT HARDWARE

2.1 Mechanical design

Most of the similar robots developed before, using the same architecture--two legs and two wheels, can only move using hybrid locomotion mode (Muscato & Nunnari, 1999; Suwannasit & Laksanacharoen, 2004). Thanks to its two legs with 2 d.o.f. each and passive wheels as feet in the front, and two actuated rear wheels, respectively, the robot developed by our laboratory (see Fig. 1) can move using three locomotion modes:

* Wheeled locomotion mode, using the rear actuated wheels (2 and 11) and the two passive wheels in the feet (6 and 10);

* The first hybrid locomotion mode, using active front legs (5 and 7) and passive rear wheels;

* The second hybrid locomotion mode, using the legs and active rear wheels.

The rear wheels (2 and 11) can be free or actuated, thanks to the two electromagnetic clutches that realize the connection between them and their servo (12 and 13). In order to offer a continuous rotational motion, the servos used for rear wheels actuation have been modified.

[FIGURE 1 OMITTED]

The advantage of our robot configuration, comparing to other similar previous designs, is an improved stability, when only one leg is in support phase. In wheeled locomotion mode, the actuated wheels (the big ones) act in the front of the robot and the small wheels, as feet of the passive legs, are passive. One of the small passive wheels or both of them (depending on the instantaneously trajectory of the robot) will be in contact with the ground, only to keep the stability of the robot. We will have two small passive wheels on the ground for a straight or curved trajectory, and a single passive wheel in contact with the ground for turning on the spot. If we consider the wheeled locomotion mode, we can write:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

and

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)

where: v--the linear velocity of the robot body; [OMEGA]-its angular velocity; R--radius of the rear wheels; [[omega].sub.r]--angular velocity of the right rear wheel; [[omega].sub.l]--angular velocity of the left rear wheel; [v.sub.r]--linear velocity of the right rear wheel; [v.sub.l]--linear velocity of the left rear wheel. For hybrid locomotion mode next relation should be respected:

[[omega].sub.med] x r = [[omega].sub.leg] x [l.sub.3], (4)

here

[[omega].sub.med] = [[omega].sub.r] + [[omega].sub.l]/2] (5)

and [l.sub.3] is the horizontal projection of the leg. In order to simplify the control algorithm, the robot has been designed as well as

[l.sub.3] = R. (6)

2.2 Control

The control board of the first hybrid robot is based on a PIC family micro-controller (Fig. 2). There are not sensors used for the first prototype that has been built only to test the architecture and locomotion modes. In order to save energy, for flat terrain only the rear wheels or the two front legs are actuated. For climbing obstacles or moving on soft or unstructured terrain, the rear wheels and the front legs are simultaneously actuated, in order to improve the traction force of the vehicle. Because the trajectories of the legs are crossing in the support phases, few precise rules should be established for the case when the robot is using hybrid locomotion mode:

--The legs could not be simultaneous in transfer phase (in that case the robot will fall down);

--The legs could not be simultaneously in support phase otherwise they will cross each other;

--When a leg is in support phase the other one should be in transfer phase, moving in opposite direction.

[FIGURE 2 OMITTED]

The testing gave a qualitative view of the system's mobility performance. All the locomotion modes have been tested, using: forward, backward, turning right on the spot and turning left on the spot sequences.

We did not yet test a curved trajectory. All the tests have been done in the laboratory. We should also test the robot on a soft ground.

3. CONCLUSION

One of the merits of walking vehicles is their superior terrain adaptability and also stepping over obstacles without applying any forces on them.

On the other hand, they are moving with smaller speed comparing to traditional wheeled platforms, which also provide sufficient robustness, mechanical simplicity and energetic performance.

One way of improving the drawback of walking robots may be to develop hybrid, in order to exploit merits of both walking and wheeled systems.

Thanks to its two legs with 2 d.o.f. in the front and two passive wheels as feet and two actuated rear wheels, the robot proposed by the authors can move using three locomotion modes: wheeled mode and two hybrid locomotion modes.

4. REFERENCES

Adachi, H., Koyachi, N., Arai, T., Shimizu, A. & Nogami, Y. (1999). Mechanism and Control of a Leg-Wheel Hybrid Mobile Robot, Proceedings of IEEE/RSJ International Conference on Robotics and Automation, pp. 1792-1797.

Guihard, M., Gorce, P. & Fontaine, J.G. (1995). Sapphyr: Legs to pull a wheel structure, Proceedings of IEEE International Conference on Robotics and Automation, Nagoya, Japan, pp. 1303-1308.

Kar, D.C. (2003). Design of Statically Stable alking Robot: A Review. Journal of Robotic Systems 20(11), pp. 671-686.

Muscato, G. & Nunnari, G. (1999). Legs or wheels? Wheeleg a hybrid solution. Proceedings of the 1st International Conference on Climbing and Walking Robots--CLAWAR-99 (editors G.S. Virk, M. Randall, and D. Howard), Portsmouth, UK. Professional Engineering Publishing, pp. 173-180.

Suwannasit, K. & Laksanacharoen, S. (2004a). A BIO-Inspired Hybrid Leg-Wheel robot. TENCON 2004, (2004 IEEE Region 10 Conference), Proceedings Analog and Digital Techniques in Electrical Engineering, 21-24 Nov., Chiang Mai, Thailand.