Robot Sheep Shearing

A Retrospective Exhibition of Pictures

James Trevelyan

School of Mechanical Engineering
The University of Western Australia

This photograph records the first shearing blow on a live sheep by a robot in July 1979. The ORACLE robot performed this feat. It was designed in early 1978 and was in service 18 months later! The shearing speed was modest by later standards - 0.05 cm per second. However, the shearing quality was excellent. Click photo for a full size picture.

A drawing of the ORACLE robot. Click for full size picture

ORACLE robot in a shearing test in June 1981. This photograph was one of many taken for the National Geographic Magazine article which appeared in May 1992. A younger James Trevelyan and Roy Leslie are watching in the background. Click picture for a full size version.

The ORACLE robot had 8 actuated joints, all with hydraulic servos.

The sheep is being supported on a temporary restraint fixture which served as one of many concept demonstrators for the ARAMP manipulator below.

The ARAMP sheep manipulator, 1984. A sheep is being shorn with the ORACLE robot during trials to demonstrate obstacle avoidance techniques. Peter Kovesi is watching. In the end, we concluded that obstacle avoidance was no real solution. The obstacles had to be removed instead!

The ARAMP manipulator was designed to support a sheep and manipulate it into the shearing positions required by the ORACLE robot. Thus it was slow and extremely complicated with 43 separate actuated movements, many of which potentially caused collisions! However, it demonstrated that sheep could be manipulated and turned over completely automatically with satisfactory reliability. It was designed by David Elford, Zbigniew Lambert and Jan Baranski with ideas from Roy Leslie.

Here is the ARAMP manipulator holding a sheep for ORACLE. Peter Kovesi is in the background. Behind him is the first full-scale model of the SM robot, still to come.

ARAMP showed that we could shear most of the sheep in a fully automatic sequence. In spite of its complexity, ARAMP was amazingly reliable. Some people were frightened when its giant jaws closed around the side of the sheep, but the sheep were never harmed.

The next step was to design a really simple, reliable robot. This was no easy task. The result was SM (Shear magic).

To do this, we had to invent a new robot wrist mechanism. We called this ET (Elephant's Trunk) because it behaves in a similar way, and looks a bit like one when it has the cover on. Here is a drawing.

Here are some pictures of the SM robot with the SLAMP manipulator. The manipulator also had to be simple, and that was even more of a challenge than the robot itself.

Darryl Cole (left) and Peter Kovesi start by loading a sheep onto this trolley. They secure the legs and head to clamps. The top part of the trolley will roll off onto a fixed platform - SLAMP.

Darryl Cole securing the head clamp.

The sheep, on its trolley, has been rolled onto the platform. Darryl Cole is checking the cutter. James Trevelyan, seated at the control console in the background, watches and will start the robot when it is ready.

The belly has now been shorn, and now SM is shearing the front leg.

James Trevelyan leans over to watch. A system of light beams prevents anyone from getting too close, otherwise the robot stops immediately. The sheep is part shorn here.

Now the sheep is almost finished. It will be rolled over and the fleece will peel onto the conveyor belt to the left in this picture.

Well, that was 1989. Much has happened since.

Shearing experiments with the robot continued until 1993 by which time we showed that high speed shearing of the whole sheep was possible. We completely sheared more than 200 sheep to iron out reliability problems.

But.....

Just as we finished the Australian Wool Industry hit a huge financial crisis. Neither the money nor confidence to develop robots could be found. The Wool industry had a debt of nearly 4 billion Australian dollars, offset by a stockpile of unsold wool which was still there four years later.

Meanwhile.......

The SLAMP manipulator now being commercialized by SLAMP Ltd and Automated Systems Innovation Pty Ltd. Darryl Cole, professional shearer and now Automated Systems Innovation Marketing Director, is shearing the neck of a sheep.

This machine is a hand shearing system based on the manipulator SLAMP designed for the SM (Shear Magic) robot in 1988/89. It has been simplified and refined for hand shearing. A semi-automatic sheep loading machine (not shown) places the sheep in the cradle, and the shearer attaches the sheep to head and leg clamps. The entire cycle including sheep loading takes between 3.5 and 4 minutes on average. A typical average shearing time using the traditional method is about 4 minutes.The shearer steps on a foot switch to make the machine turn the sheep over and move it to the next shearing position - the shearer can continue shearing while this is done.

The machine accepts a full range of sheep sizes from lambs to rams.

Compare its simplicity (4 actuators controlled by a PLC) to ARAMP (43 actuators controlled by a real-time special purpose computer system).

With this machine, shearers will be able to look forward to a full working life (average now is 5 years) without stress injuries which most shearers accept today. Shearing ability will no longer depend on strength and stamina and skill levels are expected to improve significantly with longer working periods and less physical stress. Fleeces are skirted with the free hand, saving shed labour.

May 2004 Update

As with many inventions there are many slips between the cup and the lip, as the saying goes: taking an invention to commercial reality is a hazardous process at the best of times.

The wool industry licensed the technology to a startup company SLAMP Ltd under the leadership of several experienced great years, experienced in both engineering innovation and the wool industry. Unfortunately the company was unable to raise the necessary share capital and much of the engineering work to develop the manipulator went unpaid. After unsuccessful attempts to raise capital, one of the directors bought out his colleagues shares but unfortunately he died in 2002 before he could take it any further. For a while it looked like nothing was going to happen. Then, in late 2003, the wool industry took up the issue once again and commissioned a consultant to look once again at mechanised wool harvesting and review changes in robotics and automation technology that might make the whole venture more attractive commercially. We are still awaiting the outcome of that report.

|Back to Mechanical Home Page| UWA Home page|
Web site design: Peter Murphy pmurphy@uniwa.uwa.edu.au
LAST UPDATE: May 2004