ORACLE - Our First RobotORACLE - Our First Robot
On 25th July 1979, we were once again prepared for trial number 1. A sheep arrived, trussed up with three legs tied together to prevent it from jumping out of the back of Roy's aging utility truck. It was carried in and untied, firmly held between Roy's legs, and then lifted up by willing helpers, high onto the restraint bars. It rolled alarmingly to one side, was righted, turned its head curiously, this way and that. The legs and neck were too long, and the clamps had to be readjusted. Finally the animal was immobile, panting gently, patiently.
(Stewart Key - standing - and Ian Hamilton watch as Roy Leslie - behind the cradle - adjusts the front leg clamp position with a hammer.)
Roy had shorn most of the wool from the rear end and far side of the sheep so that we would have a better view, and the robot could land for the first time on shorn skin.
The robot was slowly raised and moved across to the sheep under manual control. When the cutters were just above the shorn patch at the rear end, the position was recorded by the computer. The robot backed away automatically, and we all took a deep breath. This was it.
We exchanged glances and a silent 'let's go' passed round the ring of expectant faces. I typed the command to start the automatic approach to the sheep, and looked around for one last check. My finger hovered over the RETURN key for a second or two while I glanced at Roy who was holding the manual control box. Rob Greenhalgh held a second emergency stop button in his hand, standing well back, but ready to act. Stewart had his head down behind the video camera. My finger tapped the key, and the cutters moved almost imperceptibly towards the sheep. The cutter motor came to life, a shrill whine above the loud humming of the pump, and then there was a gentle nibbling, and rocking as the sensors reacted to the skin for the first time. A second later and the robot was moving away again. A black smear of dirty oil marked the cutter's path across the skin - the sheep seemed quite unaware of what had happened.
(The first blow - click picture for a large version)
We all broke out in smiles - but the real test was still to come. The robot was moved manually once more to mark a new landing point which would take the cutter into unshorn wool for the first time. I typed the start command, and once again the robot was slowly closing on the sheep. Again the gentle nibbling, this time for real as the cutter seemed to nibble and chew the wool off. As it pitched back and forth, the wool worked its way up into the wrist linkage where it was squeezed before falling away on the underside.
(The second blow)
A cheer broke out as we glimpsed a perfectly shorn strip left by the robot for the first time. Another blow was requested, alongside, with the same result.
We moved in and carefully collected the wool, dividing it among us - our first robot shorn wool. I still have some on my desk, alongside one of the still photos taken that day.
The sheep was taken down and trussed up once more to be returned to its paddock outside the animal house.
We were on our way, just a few months late, but elated to have achieved
such a perfect first result.
|Tour Page|James Trevelyan's
Page| UWA Home page|
Web site design: James Trevelyan jamest@mech.uwa.edu.au
September 1997