Step by Step Approach to Digital Kaizen

At the recent online ARC Industry Forum Asia, titled Accelerating Digital Transformation in a Post-COVID World, Mitsubishi Electric participated as a Platinum Sponsor. Hajime Sugiyama, Digital Manufacturing Architect, Factory Automation Systems Group at Mitsubishi Electric, spoke about digital manufacturing by providing real-life examples, with emphasis on SMKL (Smart Manufacturing Kaizen Level) and KPIs. SMKL is an evaluation index of the level of data utilization in factories. It can help management to make investment decisions by showing the purpose, impact, and relationship of each investment step towards making factories smarter.

Established in 1921, Mitsubishi Electric, a multinational electronics and electrical equipment manufacturing company headquartered in Tokyo, Japan has been continuously growing for the last 100 years. The corporate commitment “Changes for the Better,” embodies the Kaizen philosophy – a continuous process of improvement that never ends, in fact “always striving to achieve something better.” Mitsubishi Electric ranks among the top 100 digital companies globally.

This blog captures the key points of Mr. Sugiyama’s presentation. His presentation video (Day 2, Session 3) will be available at the vFairs platform for the next three months: ARC Asia Forum 2021 (vfairs.com). The documents and video link are also available at Mitsubishi Electric’s booth.

Reality Check and Digital Kaizen

Mr. Sugiyama’s presentation revolved around “Digital Kaizen,” the pain points of IoT implementations, and the lessons learned on the path of digitalization. According to a global industry survey, although 85 percent of organizations/businesses understand the potential of Industry 4.0, only 10-15 percent have detailed strategies in place. The reason for this is because there’s a gap between the vision/dream and the reality on the shop floor, explained Mr. Sugiyama. He elaborated on this through the struggles faced by their customers. When a new production line is introduced, it seems like a good opportunity to implement Industry 4.0 (full automation, AI/machine learning, digital twin, cloud, 5G and so on); but the proposal is usually rejected because of the high cost of implementation and the difficulty in calculating return on investment (ROI). Hence, the management decides to wait until they have a better grasp of the technologies and costs can be reduced.

Mr. Sugiyama gave the example of a data analytics project for predictive maintenance of a plastic injection machine to improve output quality. This was done by real-time data collection and analytics deploying edge computing and AI. It succeeded to a certain point, but the project was postponed as there was a low failure prediction rate, and the root cause of the problem could not be pinpointed. More data was analyzed, but the outcome didn’t improve. The company’s best engineers were on this project for nine months and the management believes that it was a waste of valuable resources.  Mr. Sugiyama advised, “Use your time wisely and focus on areas where you can be more productive.”

Other causes for failures in IoT Implementations in factories:

• High consultancy cost
• Energy monitored, but savings not realized
• High installation cost due to old equipment
• Complicated data analytics

In-house Motor Assembly

Using the example of a motor assembly from one of Mitsubishi Electric’s own factories, Mr.  Sugiyama summarized the following:

Pain points: Long training period; assembly errors; difficult to assess efficiency.

Digital Solution: Installed a Smart Work Navigator, and this reduced the training period, eliminated assembly errors, automatic measurement of assembly time, and data was checked while tightening screws. This digital solution was very useful when social distance had to be maintained during the pandemic, and employees were brought up to speed in a short span of time. All these factors helped achieve return on investment.

Next, Mr. Sugiyama highlighted a few challenges of this solution, such as the time taken to program it, “kitting” (collecting the parts required for each product) of the screws can be time consuming, and some of the data may not be utilized.  To overcome these bottlenecks, a three-pronged improvement process was put in place: simplify the preparation of the digital manual, select screws at the assembly table, and introduce a data utilization system. Although the ROI period was long, the system could be reused for other production lines. 

Lessons Learned

It is important to dream big, but start small, explained Mr. Sugiyama. If the scope is too vast then it will be difficult to calculate the return on investment, but if the scope is smaller, the ROI can be predicted. Sharing and agreeing on the goals and KPIs (i.e. where you are, at what level you are at, and the direction you want to go) with the management is another important aspect. 

Mitsubishi Electric uses the SMKL matrix that evaluates the level of digital manufacturing by using two axes: one is the maturity level, and the other is the management level. The level of data visibility is represented by the maturity level, shown on the vertical axis. The horizontal axis represents the granularity, or level of detail for the items to be managed.

Further, Mr. Sugiyama said that IoT must be implemented step by step (collect, visualize, analyze, and optimize). It’s the same at the factory level, you have to begin small and then move up. At Mitsubishi Electric, an SMKL map is charted to discuss where they are and where they hope to be in the next five years. It also helps to map the staff’s skill and knowledge levels. Mitsubishi Electric is collaborating with other companies in Japan to standardize the SMKL and KPIs approach.