Today industrial and manufacturing organization are using equipment, machines and systems which generate large set of data. With the advent of technology and availability of tools, companies want to gather more and more data about their processes, products and people. It has been observed that not all the data gathered is applied and analyzed to develop meaningful and actionable insights. As per a research report, more than 70% of the data which was generated on shop floor is never consumed.
Question remains, how much is too much and how to know what data I should capture in the early stages of an IIoT project.
Delivering a Lean IIoT project
In the later half of twentieth century, when the manufacturing gained prominence and automotive companies pioneered the new ways of manufacturing, 5S become a way of managing shop floor for Japanese organizations. At high level, the 5 steps in 5S focuses on removing the unwanted, prioritizing the wanted, cleaning them, standardizing the practices around them and making sure that everyone else follows those practices.Â
It remains a highly useful tool for manufacturing organizations even today but as the organizations move from traditional operational excellence practices to digital and technology based tools, the question arises, how relevant will 5S be?
The concepts of Lean, including 5S remain relevant, in-fact these concepts are more crucial in digital era as the stakes are high. If we take 5S, it could be directly applied in optimizing the execution of an IIoT project. The concept of 5S in digital could be redefined as ‘application for every data and every data in an application’.
Let us explore how 5S is applied in execution of a digital project:
1S – Sort: The first step is to remove all the unwanted items from the system. As the project has clear goals aligned with KPIs, team should evaluate that all the data that is being captured is applied and analyzed. In addition to that it is also important to check the frequency of the data capturing and its implication during analysis. If a data pointer will be used only on hourly basis, capturing it on minute basis is adding waste and cost to the process. There are some cases where equipment data needs to maintain in Process Control System (PCS) for troubleshooting purposes, but not necessary for capturing in IoT platform.
As a outcome of this step a list of data pointers with sources to be generated and data pointers should be categorized as – wanted in the project, wanted but not now and not wanted.
2S – Set in Order: Once you have clear list of operational KPIs needed to be calculated and data that needed to be tracked, prioritize that list and map each data pointer to an application and also to the KPI. The parameters which are displayed post analysis, are linked to specific KPIs and will help in generating actionable insights.
As an outcome of this step, all the data pointers are mapped to specific application and every application is linked to operational KPIs.
3S – Shine: The 3S focuses on the upkeep of all the hardware and visual appeal of your applications including dashboards, reports, notifications and alarms. All the hardware that is being used for data collection should be properly identified and marked for ease of accessibility and maintenance. The applications should be customized as per function and persona. Users at different levels across functions should only see the data which is relevant for their roles. This will also help keeping all the applications like dashboards, reports and alerts clean with optimum data. Too much data and analytics at application level not only makes it difficult for the user to read and understand, it also distracts the focus from key points.
As an outcome of this step, the user requirement at persona level for each function should be mapped and data to be analyzed and projected to be listed.
4S – Standardize: As the organization starts expanding the horizon of IIoT projects to add more lines, processes and factories, it is crucial to define the standards. The basic procedures should be defined across the organizations for the hardware, network, firewalls, softwares and IIoT platform. These procedures should become guidelines while initiating an IIoT project. All the teams and business units should adhere to these procedures and ensure that they are installing the compatible hardware, software and IIoT platform. A key aspect to check here is that, all your systems should talk to each other without any additional customization or need for middleware.
As an outcome of this step, an IIoT execution procedure and 5S audit sheet should be developed and shared with relevant stakeholders. Â
5S – Sustain: For success of any initiative, it is critical that the procedures which are put in place are followed. A dedicated team should be formed and all IIoT projects should be evaluated on the aspects mentioned above before execution begins. Team members should be trained to evaluate the projects and also keep a tab that project adheres to basic guidelines. A 5S team should be formed and roles should be defined.
As an outcome of this step, each IIoT project should be perdiodically audited as per 5S audit sheet and report should be presented to management.
Conclusion
In an IIoT project, for every extra byte of data that a company captures, stores and applies it pays a cost and if that does not generate any actionable insight that data is actually waste. By capturing colossal data volume, there is a possibility of missing the right set of data required for business use. It takes a lot of time to find and sanitize the data and then generate actionable insight. Such data distracts organizations focus, adds to cost and also impacts the environment at large.
It is very crucial for the companies to clearly identify the business problems they want to solve, select relevant operational KPIs, list down the data pointers which are needed to calculate those KPIs and prioritize to capture those data pointers.
Every data that a company captures, should be analyzed, applied and generate actionable insights for business improvements. If not so, do you really need that?
Every data, in the end, should help its users to take meaningful action.
The article is co-authored by Suman Chakroborty, Consultant – Intelligent Manufacturing, Hitachi Vantara, and Ankur Chaudhary, Manager – Manufacturing Consulting, Hitachi Vantara.