"NOT in my life time"

Shearing jobs look safe from robots

for some time to come

by Sharon K. Parker

 

Fears of digitalization taking over jobs have ramped up, with many commentators predicting massive job losses in the future.

 

Yet it seems unlikely that robots will successfully shear sheep any time soon. “Not in my life-time”, predicts Cartright, a world-record-holding shearer. “Robots can do it (shear the sheep)”, he says, “but they cost too much”. Bill, a 67 year old veteran shearer agrees, pointing to the corner of the shed:

 

“Right there – that’s where they tried a machine set up shearing sheep. It was useless, it couldn’t keep up with the blokes”.

 

Failed attempts at sheep shearing machines

 

There have certainly been attempts to shear sheep with robot-like machines since the late 1970s[1]. But, as Steve the boss observed, “they spent $10 million on trying to do that and the machine is sitting over east in someone’s paddock.

 

So shearing robots have never been very successful. Why is this so?

 

Shearing is an excellent example of work that is high in tacit knowledge. Tacit knowledge is ‘knowing how’ to do something, yet not being able to clearly articulate or write down this knowledge[2]. At least historically, because of the difficulty of articulation, it is difficult to acquire and transfer tacit knowledge from skilled human workers to robots. As one of the shearers explained:

 

 “It (shearing) is a feel job… A robot can’t feel. You feel the handpick, you feel the sheep, and the sheep feels you and knows your positions… it’s a whole body job.”

 

Robots can't do the whole job

 

With machine learning, robots are getting cleverer at taking over those tasks we previously assumed were too tacit for machines to learn. But there is still a challenge. Shearing highlights a crucial point about digitalization. And that is that robots and machines are much less likely to cause change in whole jobs, but rather will affect tasks within jobs.

 

So even if there was a robot shearer that was faster and more cost effective than a human, and that could cope with all the ways that sheep vary, someone still needs to catch the sheep and get it locked in the right position. And someone needs to put the sheep back after its been shorn. And someone needs to pick up the wool, and decide its quality. The massive investment needed to do all these different tasks using robots, in a co-ordinated way, is unlikely to ever justify the savings. It is simpler and more efficient for humans to just shear the sheep. So, in the end, Cartright’s prediction of “not in my lifetime” seems spot on.

 

Help people instead of replacing them

Instead of trying to replace humans with machines - the focus of many technological efforts - perhaps the goal should be to use technology to help humans to do the work. Maybe, for instance, effort is best spent improving the way shearing is done[3] to better protect shearers’ backs from injury[4].

 

References

[1] See, for example, https://www.youtube.com/watch?v=6ZAh2zv7TMM

[2] Polyani,M. (1958) Personal knowledge. Tacit knowledge

[3] See, for example, http://blog.mecardo.com.au/an-evolution-in-sheep-shearing

[4] Gregory, D. E., Milosavljevic, S., & Callaghan, J. P. (2006). Quantifying low back peak and cumulative loads in open and senior sheep shearers in New Zealand: examining the effects of a trunk harness. Ergonomics, 49(10), 968-981.