Ask yourself a question: “Will the way that we manage and improve our processes now change in the next five years or will we still be using the same tools and approach?”.

Profound changes are already being implemented in manufacturing and other fields (Industry 4.0, Internet of Things (IoT), advanced automation, AGVs, Artificial Intelligence). To get the most out of these opportunities we must leave some of the old hallmarks of process improvement behind and shift focus to better tools.

Five years ago I was using OEE analysis to guide my process improvement projects. Now I use process simulation software, here I’ll summarize what I have learned and how I benefited from making this transition.

OEE; what is it good for?

First introduced in the 1960s and popularized as part of TPM (Total Productive Maintenance), OEE or ‘Overall Equipment Effectiveness’ is a performance metric that indicates the productivity of an activity or Line by assessing 3 ratios:

• Availability
  Uptime, or the amount of time spent working versus the planned production time. This score diminishes if stoppages, breakdowns or changeovers take away from the amount of time that could have been spent producing output.
• Performance
  Actual throughput versus the maximum throughput that could have been produced in the available uptime. This score diminishes if longer than ideal cycle times occur.
• Quality
  Total parts processed versus number of parts produced right first time. Likewise this score diminishes as more bad parts or more reworks occur.The OEE score is the product of these ratios:OEE = Availability × Performance × Quality

Since then, OEE has retained great popularity and has been adopted by many industries as a favorite performance metric. Like me, many of you will have been introduced to OEE as both a scorecard and a process improvement planning tool. In this capacity, there are certainly practical benefits to using OEE:

• Universal application
  It can be used to assess almost any output based system. This allows benchmarking to occur between vastly different plants and operations.
• Easy to calculate
  Providing you have collected the data, defining the OEE of a given activity or station is straightforward formula work.

What do we gain by moving our focus beyond OEE?

OEE is a good monitoring tool, however its limitations become apparent when we try to extend its functions into planning and managing process improvement. ‘Constraint identification’ is a good example of this. OEE can be used to identify the most under-performing activity in a process. However, this approach has a critical flaw – it does not consider item flow. Without this we cannot see how an activity fits into the overall process and we cannot then predict the impact of targeting the activity for process improvement.

For example, what happens if the under-performing activity exists within a group of machines that collectively are actually over-performing? This would negate the benefit. Or as another example, what if solving the constraint only functions to hurry items into another, previously hidden, bottleneck? Again, there would be no gain.

I realized that to manage my projects more effectively I had to move beyond OEE (i.e. put it in the trash!) and use a process improvement tool that could demonstrate item flow and also allow experimentation to predict performance impact across the whole process (a key element for intergraded systems such as Industry 4.0 and IoT).

This is why I made the transition to using Simul8 process simulation software. In comparison to using OEE as a guide for process improvement, there are a range of key benefits I’ve found from using simulation.

What are the benefits of using simulation to support process improvement decision making?

• Understanding item flow and the impact of changes
  Tracking the movement of individual items, products or parts gives us a much more useful set of metrics for predicting system performance and throughput. Understanding the lead times and queueing times of items is a better indicator of bottlenecks and allows us to test a wider range of improvement options such as specialized item routing, single piece flow, push versus pull, Kanban placement and other Lean or automation strategies.
• Scenario testing before implementation
  If you are considering making just five changes to a production line (with each change possible at three levels of intensity) this already creates 243 possible scenarios of how that line will perform. This is a considerable amount of data to manage and make sense of. New technologies push this further as automation and AI allow our processes to become much more agile. The nature of these investments also cause more of the process design to be locked-in at an earlier stage of the procurement process: Did you correctly predict the max size limits on parts, or the number of AGV’s required in the system?Making corrective changes to these types of parameters post-implementation is becoming an increasingly costly endeavor as we trend towards smarter and more expensive machines. Why take the risk when for a fraction of that cost we can test all scenarios well ahead of implantation? Take that one step further; what are the extra benefits that having a simulation of the process has when it comes to designing that AI or those AGV paths in the first place? Increasingly we are seeing simulation give direct insight to system designers. This can be in the form of routing rules, priority logic, testing of max system capacities or even PCB logic. These are all elements that I have seen be literally copied and pasted from Simul8 models into the parameter criteria of emerging process automation / AI systems; speeding up their development and ensuring a correct fit when the real-world system goes live.
• Visualize the change and get buy-in
  Obtaining support for your project from either a client or senior staff is a massive part of delivering a successful process improvement project. Even the most brilliantly devised process change will get nowhere without the buy-in of key decision makers. In my experience, presenting changes in the context of positive OEE metrics or similar chart or formula based proposals does not do enough to inspire and inform decision makers on what you are trying to achieve.Using a simulation to present your process improvement concepts is a vastly different experience. By animating the process flow the audience can instantly gain an understanding of how the system is performing now and where your changes will impact this. We can follow a single item through each stage of the process to show its progression against target times or link a group of similar items together by color to quickly highlight where they face most delay in the process. As an interactive tool, simulation also allows for audience engagement: “but what if that process takes longer than predicted?”, “could we outsource those stages?”, “will we need a back shift to support that throughput target?”; these are typical of the sort of questions that we can instantly assess by editing the simulations input values, then re-running to show the impact.

Making this type of system-wide and investigative analysis is simply not possible when focusing on an OEE based approach. As new technology opens up a much wider range of possibilities for how we manage and run our processes the demands and rewards of process improvement shift to tools that can facilitate bigger data sets and allow for experimentation.

So is OEE set for the trash?
I would suspect not. Established metrics like OEE will always aid our understanding of our processes and stay relevant but broader and more visual approaches such a process simulation are better suited to get the most out of ongoing trends and represent a natural progression to the types of tools we use. Like the transition from driving with maps to driving with GPS; getting around by sketching out a fixed route on paper has now developed and switched over to a more visual, focused and interactive platform.