Those with some experience in lean manufacturing have probably heard the phrase from Toyota’s Taiichi Ohno: “Data is, of course, important in manufacturing, but I place the greatest emphasis on facts.” What he suggests is to not only trust reports but also to periodically walk the gemba – a Japanese term that means working environment.
When discussing energy management practices in different companies, we often hear “proud” managers arguing that they are doing very well because they have thousands of tags being registered in a historian system, they already receive a daily consumption report for each production line, they account for all costs on the first day of the following month, or even that they have the history of the last 3 years of invoices for the thousands of sites in their operation. It doesn’t take long – usually around 5 questions – for them to understand that all they have is data, and because data alone does not carry information from the gemba, it does not reveal the relevant facts that would lead to effective actions.
The objective of this post is not to elaborate on the fictitious hierarchy of “data, information, and knowledge,” but to demonstrate how an adequate energy management system can promote fact-based management.
It is the job of an energy management system to keep a history of relevant measurements from consumption meters, different types of instrumentation, and human observations. Isolated, this data helps answer questions such as: What? Where? When? How much? We can, therefore, report how much each piece of equipment has consumed of each energy input in a given period of time. Even without putting into question whether the data is correct or not (this will be the subject of an upcoming article on the quality of energy data), we may criticize this approach solely based on consumption data as a crucial question remains: Why? “Why did we see that specific behavior from that particular asset?”
The missing ingredient in the above method is what we call contextualization. Only from knowledge of the context of the equipment (from the gemba) can we explain the reasons behind its behavior. Abnormally high consumption may be explained by factors such as external temperature, type of product being made, equipment setup, raw materials, operator responsibility, volume produced, among others. These attributes give context to consumption information, helping explain them and highlighting relevant facts.
A good energy management system aligns measurement data with contextual information over time, creating the conditions for analyses of asset behavior. It allows breaking down consumption KPIs over contextual variables. It monitors the real performance of each asset and benchmarks it to the behavior expected for the current context at every moment.
Having a measurement infrastructure that collects large amounts of data is a necessary but not sufficient condition for effective management. Perspective, context, reference, and understanding are equally necessary. Only through this integrated approach is it possible to manage based on relevant events. Managers selecting an energy management system to assist in the quest for greater operational efficiency must ask themselves whether data and reports are enough, or if they should “place the greatest emphasis on facts.”
Viridis builds energy & utilities management solutions based on big data. To us, however, large amounts of data are not enough. Our products employ contextualization algorithms, pattern recognition algorithms, automated model inference, and relevant fact identification from a large data lake fed from sources all over the operation. One of our differentiators is to help answer the “whys” of energy consumption and effectively promote actions to improve efficiency.