Written by Malcolm Jones
Industry 4.0 is being hailed as the Fourth Industrial Revolution, although some might say that the introduction of Lean processes by Toyota in the 1950’s and onwards should also be counted as an additional industrial revolution. There are strong links between what is now happening in Industry 4.0 and the development of Lean.
If we go back to the beginnings of Lean in the 1930’s, Toyota based their process on two pillars, JIT and Jidoka. JIT meant producing what the customer wanted, precisely when they wanted it and Jidoka means separating people from the equipment or process. The first patented Jidoka device was the tensioner on the Toyota Automatic Loom which stopped the loom whenever a thread broke, removing the need for a separate worker to watch each loom to manually stop it. This allowed one person to manage a large number of automatic looms. Note that one function of Jidoka is that the machine tells the worker that there is an issue which needs addressing.
JIT also relies heavily on good demand information over extended supply chains. John Sterman and the systems group at MIT have demonstrated the effect of interruptions to the information flow in lean systems, the demand amplification or ‘Bullwhip’ effect where demand is distorted by information lags, generating excess inventories as well as stockouts, the antithesis of Lean, and our lean approaches attempt to mitigate this by synchronising information and material through Kanban techniques.
This is where Industry 4.0 meets Lean, in the provision and management of information. Industry 4.0 is based on the digitisation of information and its transmission through web based applications. Instead of using semi manual Kanban and Heijunka systems to synchronise information and material flow for production scheduling, technologies such as RF tags can be used to assign material to workstations and manage flow through the factory in real time.
In this article however I am more interested in the implications of Industry 4.0 for the management of Lean Equipment through Total Productivity Maintenance (TPM), as often used in a lean system. TPM is usually structured according to pillars of activity, including Autonomous Maintenance for the production group, Planned Maintenance for the maintenance group and Quality Maintenance for the quality group. All these activities are inherently integrated, but for some purposes it is useful to consider them separately.
The objectives of Planned Maintenance are to maintain the optimum reliability of equipment at optimum cost. To do this we rank the criticality of equipment to the production process and perform FMECA (Failure Mode Effect and Criticality Analysis) to identify the optimum maintenance approach to specific items of equipment.
One concept in common use is the P-F curve from reliability engineering. This curve is used to illustrate how potential (P) of a functional failure (F) can be detected and mitigated before failure occurs.
This is based on the observation that most failure is statistically random – there is a constant probability of failure over time, so that time based PM schedules are inherently inefficient, either under maintaining or over maintaining. The time interval which is of most interest to reliability engineers is the P-F interval, the interval between being able to detect potential failure and the failure occurring.
As the above graph illustrates, condition monitoring technology is far more effective at detecting potential failure than manual inspection, and in particular can detect potential failure far earlier, allowing preventive maintenance activity to be scheduled to avoid functional failure.
Based on our FMECA we would normally set intervals for inspection using particular technology, often, but not necessarily, conducted by external specialists and also requiring expert interpretation of results. The goal of TPM and reliability engineering has always been to move from reactive to planned to condition based maintenance, but the barrier has been the cost of the technology and the expertise required for its use. The promise of Industry 4.0 is that web based applications can use real time data from machine monitoring systems and even expert systems for interpretation to provide continuous feedback on critical equipment, making the TPM goal of Zero Breakdowns an achievable reality.
Some of this is futurology, particularly in terms of expert systems which can interpret condition monitoring data, but much of the condition monitoring technology and data gathering capability already exists. The implications for the maintenance department are that in the medium term considerable upskilling will be required in terms of the implementation and interpretation of condition monitoring data.
Autonomous Maintenance is a distinctive feature of the TPM approach to Lean equipment management. Autonomous Maintenance has two fundamental objectives, firstly the restoration of equipment to optimum condition and development of procedures to maintain that condition, and secondly to build the production group’s skills to the point where they are autonomously managing the equipment, only calling on outside resources for specialist maintenance tasks.
Two features of this approach are the development of inspection standards and the use of an abnormality tagging process to identify, record and correct abnormalities. One common issue is that the paper based Autonomous Maintenance procedures need to be duplicated into CMMS systems to provide a searchable database of equipment abnormalities which can then be used to optimise equipment reliability.
The promise and use of Industry 4.0 in Autonomous Maintenance is threefold, all based on the use of web based tablet technology. While the Autonomous Maintenance process of equipment restoration will always be a hands-on activity for the production team, supported by maintenance, web based tablet applications can enhance this. The first step AM is to gain an understanding of the equipment function. Tablet based instruction through simulations is an effective way of doing this – a simulation can show the functioning of the equipment in far more detail than observation and explanation.
Once teams have restored equipment to its optimum condition they are then tasked with creating and implementing autonomous standards to maintain that condition. Typically these would be checklists and One Point Lessons illustrating the check to be carried out. The benefit of making this tablet based rather than paper based or even workstation based is that hyperlinks can move directly from the checklist to the One Point Lesson while the operator is making the inspection, and the results can then be recorded in real time, and if quantitative can even feed the reliability database to provide predictive information.
Another use relates to the ‘tagging’ process where abnormalities are recorded by production staff on paper tags which may be attached to the equipment itself and also have duplicate copies to serve as input to the work order system. A tablet based tagging system would have several advantages, not least the feature that data is input only once into the system and can also be shown visibly on equipment diagrams, making it easier for production to describe an abnormality and for maintenance to understand the issue being raised. Electronic displays can indicate the number and location of abnormalities on a visual of the equipment, and even indicate resolution rates and leadtimes.
Quality Maintenance is the term used in TPM to denote process control activities, particularly those focused on the quality of the equipment output. In common with everything in TPM, the goal is zero losses, in this case Zero Defects. Process control is particularly important in a TPM environment which aims for zero defects on a first pass basis rather than relying on QC inspection to prevent defects reaching the customer. QM is based on a particular thought process called P-M Analysis, focusing on the physical mechanisms where product interacts with process and establishing control points with known process parameters.
The intersection with Industry 4.0 is the monitoring of that process data, together with real time process adjustment. Just as a machining centre might adjust its feed based on tool wear, we can theoretically control any process automatically once we have correctly identified the control points and parameters. Much of this is already in place in control rooms in process plants, but the next industrial revolution lies in small scale automated process control for discrete equipment.
An article such as this is not the place to advertise particular technical solutions, but one starting point for companies could be looking at their use of condition monitoring technologies, investigating current monitoring solutions and developing their capabilities in both people and technology so as to be ready for the promised advances of Industry 4.0.
Industry 4.0 is the next step in Jidoka, the separation of people from machines, where machines provide real time information to enable both the optimisation of flow through the process and the optimisation of equipment reliability and performance. I believe Sakichi Toyoda, inventor of the Toyota Automatic Loom and father of the founder of Toyota Motor, would have welcomed Industry 4.0 as the next step in his quest to develop human automation (Autonomation, another name for Jidoka), freeing workers for more value adding and creative work.