Last month, we talked about waste as ‘The Enemy of Lean’ and gave one or two examples of my own experiences during my long and continuous career supplying performance improvement systems to the food and drink industry.
Though there are many more anecdotes worthy of sharing, and I have no doubt that readers could add many of their own, it would be unfair to focus solely upon the downside and give the impression that ‘all is doom and gloom’. Far from it.
The waste may be more apparent to fresh eyes. Because it has been in evidence for so long it is often no longer seen as such, by those most familiar with; just treated as a normal part of the production/conversion process.
On the other hand, companies may well be aware of the waste and even its magnitude but lack the detailed and reliable granularity necessary to determine and fix root causes. Very much like the mid-size producer of special sauces who recently shared with us that they lose in excess of £1million pound per annum of their raw materials in scrap and wastage, but lack the detail necessary to determine root cause.
Waste, most accept, is a necessary part of the production process. However hard we try, we can never eliminate it but, in most cases, we can do much to reduce it, so long as we have sufficient visibility, analysis and prioritisation to drive us, firstly towards root cause and, secondly towards minimisation.
Automation and AI (Artificial Intelligence) can definitely help in some circumstances but this can easily be overdone. I recently visited a well-known cake manufacturer who shared with me that they had spent in excess of £2million on end of line robots to automatically pack their delicious bite-sized cakes. On visiting the line to see how this was going, we found that, during a period when 20,000 of these bite-sized goodies should have been packed, only around 4,000 had been. In asking why, the Manufacturing Manager was told ‘Dunno! Must have had some downtime’.
Automation, without real-time accurate information, analysis, prioritisation and instant visibility often serves few, if any, improvement objectives. In the above case, we do not know the detail, but our suspicion is that, prior to the implementation of robotics, senior management failed to capture the ‘hearts and minds’ so essential to ensure successful outcomes. Without this operational ‘buy-in’, suspicion and fear reign supreme. A recent study suggested that far from reducing jobs, AI, well implemented with the full backing of the operational team, should increase jobs by the year 2020.
Increasingly, at Harford we find ourselves doing much of the work often done by consultants. This is not to knock the excellent work of many consultants, nor to compete with them, but to recognise that continuous and sustained improvement must begin with ‘Why?’ and the ‘why’ must be shared with the whole operational team, such that, if the anticipated improvements (automated or otherwise) are to be effective and lasting, they must be shared with the whole team before the automation process begins.
It is no longer sufficient, if it ever was, to inflict a new system upon the workforce and expect them to embrace it. If our experience has taught us anything it is that such infliction is an easy way to fail, as the operational personnel may prefer to use their ingenuity to sabotage the new system, through fear of the unknown and concern about the loss of overtime or jobs, rather than to embrace it for its intended benefits.
Many processes, within the food and drink industry, are just not suited to full-scale automation, largely due to the short batch runs and frequent changeovers now commonplace amongst manufacturers to satisfy their supermarket customers.
‘You cannot control what you don’t measure’
In any new application, we would start with the basics. We would want to understand ‘Why’ and what could be the consequences of doing nothing. If, having done the groundwork, some processes could be cost-effectively improved by automation, we would engage appropriate experts. If, as is more often the case, some readily available tools need to be better deployed, such as SPC and other Six Sigma tools, and better measurements taken, we would begin with these.
After all, improvements to any conversion process can only deliver upon two possibilities, better raw materials and time utilisation, both of which seem to carry high levels of waste in many organisations. Accurate measurement of such parameters, at every production stage, has never been more important and, the old chestnut that “You cannot control what you don’t measure”, has never been more true. It was recently said “As technology becomes more intelligent so should we” and “If Artificial Intelligence can be used to plug a gap where we are weakest (Automation or not), then we should use it.” Such circumstances shouldn’t be seen as an opportunity to eliminate jobs but opportunities for performance improvement. If such improvements mean that some jobs are lost then that’s a pity, but more often than not we believe that improvements generally lead to an upskilling of the workforce to set up and control the AI, whilst the automation itself takes care of the boring repetitive jobs that many human beings don’t want to do anyway.