American Society for Precision Engineering
Abstract Review
ASPE Home Page Join ASPE ASPE Publications ASPE Meetings ASPE Calendar Employment About ASPE



Abstract ID   1917
Title   An In-pipe Mobile Inspection Robot Provided an Artificial Intelligence and Passable at Several T-junctions
Category   Equipment, Machines & Instruments: Mechatronics
Primary author  
Organization  

Content   We have many small diameter pipes that are gas or water pipelines for individual or corporate houses. They must be inspected in order to protect from the accident previously. Since these pipes are buried in the underground or buildings, it is very difficult to inspect the pipes from the outside. Moreover, the pipes have T-junctions at the points where the main pipe and branch connect. Therefore, the inspection robot must recognize the T-junctions by itself ad turn or pass the T-junctions. Authors have researched in-pipe robots which are driven by air pressure and can pass a Y-junction and a T-junction [1], [2]. However, they could not recognize junctions by themselves, because they have no sensor of the sight.

Now, authors fabricate a robot which has an optical sensor and an artificial intelligence and can pass T-junctions by itself. The fabricated robot is structured by a moving mechanism to move in the pipe, a steering mechanism t pass T-junctions, an optical sensing mechanism to sense T-junctions and a computer for an artificial intelligence. For the moving mechanism, we use five rubber bellows actuators which are 33 mm in outer diameter and 60 mm long. For the steering mechanism, we use four rubber bellows actuators which are 16 mm in outer diameter and 60 mm long. The optical sensor is structured by four light emission diodes and four photo transistors. The optical sensing mechanism is arranged at top of the steering mechanism. The light from the light emission diode is reflected at wall of the pipe and is received by the corresponding photo transistor. If there is the wall, a voltage of the phototransistor is 4.8 V. On the other hand, if there is no wall, a voltage of the photo transistor is 2.5 V. The optimal threshold is confirmed to be 3.5 V in case of no wall. The recognition mechanism to recognize the T-junction is a comparator. In the comparator, the voltage of the photo transistor and the threshold of 3.5 V are compared. If the voltage of the photo transistor suddenly changes from the higher value than the threshold to the lower value than the threshold, it is confirmed that there is no wall. That is to say, the robot arrives at the T-junction. Next, the computer judges it the T-junction which the robot must turn or not. If the T-junction is the T-junction to be turned, the computer orders to turn and pass the T-junction using the steering mechanism.

The fabricated robot provided an optical sensor and an artificial intelligence is confirmed to recognize T-junctions and to turn and pass to the desirable direction of the pipe.

[1] S. Kato, M. Shirakawa, M. Ono and S. Fukumoto, “Fabrication of In-pipe Mobile Micromachines Driven by Pneumatic And Vacuum Pressure”, Proc. of ASPE 1999 Annual Meeting (1999),pp37-40.
[2] S. Kato, M. Kato, S. Ogawa and M. Ono, “ Development of Inchworm type Mobile Robot Movable in Pipes with T-junction”, Proc. of ASPE 2004 Annual Meeting (2004), pp257-260.

 

Home | Member Home | Membership | Publications | Meetings | Calendar | Employment | About ASPE | RSS Feed


American Society for Precision Engineering
P.O. Box 10826, Raleigh, NC 27605-0826 USA, (919) 839-8444, (919) 839-8039 Fax
© 1999-2010 American Society for Precision Engineering. All rights reserved.