Robotics is not always the answer – Have we asked all the right questions?

by | May 12, 2017 | News

I was speaking with my colleague Chris Bennet the other day about line optimisation.

He had seen a factory with multiple production lines that was seeking to use robotics to automate the case packing and palletising processes. The issue Chris had was that the end of line was not the pressing issue. There was a bottleneck in the processing operation causing a net reduction in line packing speed of 15% of productivity at the packing end of the line. This resulted in frequent line stops and erratic start processes. What was required, argues Chris is that the bottleneck is temporarily fixed using additional labour to identify the maximum efficient speed that can be achieved manually before the application of any robotic automation solutions. This benefits the business in 3 ways:

  • The line is optimised for manual operation making a finance saving with limited investment
  • The business case for automation can focus on a solution to improve the optimised line rather than a bolt on solution which makes the best of an average job.
  • Once we know the true maximum pack line speed it may be that an alternative solution might be more efficient, cheaper or flexible for future as yet unknown needs.

I suppose that what I am saying is that applying robotics to a broken process may well provide a return on investment and make a sound business case. But it might not be the optimum solution.

A fully optimised solution really drags every penny of value from the process whilst also having the freedom to identify and cater for future requirements.

I used to be involved in software and the “paperless office” in the 80s and 90s.

(it was forecast then that there would be no paper transactions left in the office by 2000)

There used to be a phrase “garbage in, garbage out” when discussing IT automation systems. If you automate an inefficient manual process without first optimising the whole process, there is a danger that you end up with an inefficient automated process.