For many manufacturers, the path to building a Smart Factory is still confusing because of information overload. In order to overcome this challenge, manufacturers should view this transformation as a journey with four stages that reap ongoing benefits to their operations. As with any extensive company-wide transformation, trying to achieve the end goal too quickly can leave you back where you started, having wasted time and money.
It’s critical that manufacturers understand that the Smart Factory is primarily about data.
Prior to the Fourth Industrial Revolution, commonly known as Industry 4.0, manufacturers relied on clipboards and manual methods to collect machine data, perform root-cause analysis, or gain insight into their operations. But as the competitive landscape of manufacturing changed, and consumer demand increased, the industry reached a point where these manual processes were no longer efficient. In fact, they cost manufacturers time and money in the form of lost productivity, suboptimal machine output and product quality.
The Smart Factory evolution is about building upon the advancements of the Third Industrial Revolution by automating the collection of data from machines and applications, and transforming that data into immediate insights. This new technology turns the tedious, but critical, process of extracting insights from data into one that is instantaneous, streamlined, and achievable for every manufacturer.
Level One: Available Data
A level one system is pretty much the status quo. At this stage, data is available but difficult to use to make decisions or implement improvements. The data is in siloed systems, requiring manual work to integrate and translate into useful information. Problem solving at this level is possible but extremely time-consuming. When a product quality or machinery issue arises, operators and engineers must scramble to manually gather data from various systems before they can ascertain what happened and how to fix it. This manual approach is not only frustrating, but costly; it drains time, resources, and money from the factory. Manufacturers at level one should move to level two as soon as possible or risk wasting millions of dollars in lost production output from unplanned downtime each day.
Level Two: Accessible Data
A level two system integrates all the disparate data sources into one single source of truth and continuously gathers and tracks production data. With the data in one location and always available, problem solving becomes almost frictionless. When an issue occurs, operators and engineers can access the data in the system using data visualizations and dashboards—essentially leveraging the system as a query engine. With easy access to all the data, they are able to answer questions quickly, increasing plant productivity.
In addition, a level two system allows engineers to focus on addressing high-value issues such as improving the product itself, changing materials, or adopting a mass customization strategy. However, at level two, proactive analysis, which enables factories to make improvements before issues occur, still requires time, effort and engagement from engineers.
To move from level one to level two, manufacturers must implement a new data architecture, which takes only a matter of months. To do this, you need to evaluate whether to build your own system or select the right solution providers and partners. Also, when selecting a new architecture, make sure it allows you to scale the amount of data you can collect without paying higher marginal costs or sacrificing system performance.
Level Three: Active Data
A level three system shifts manufacturing operations from reactive problem solving to proactive analysis and improvements. The system enables operators and engineers to be truly preventative and proactive in solving problems, which would not be possible in a level two system.
To move from level two to level three, you must build on the previous level’s data architecture by adding new system capabilities such as machine learning and artificial intelligence. These new tools allow you to start generating insights in as little as two or three months, depending on your product mix. These new features, combined with the level two system that aggregates all your production data, create an intelligent system that on its own will find valuable insights and predict failures more accurately, while delivering information to the appropriate person at the right time. Users do not have not have to query the system or perform manual process analysis in order to find the answers to solving production issues.
An example of level three system attributes include machine learning models that predict product defects or machine failures and identify ways to produce products more efficiently. In a level three system, a person is still needed to make the changes that the intelligent system recommends.
Level Four: Action-oriented Data
At level four, the data system actually deploys the recommendations that it finds from analyzing manufacturing data. For example, a machine learning model will identify an optimization, then generate and send the recommended new settings to the machine, where it is automatically executed. In such a closed-loop artificial-intelligence-controlled production line, the time it takes to execute on an insight discovered by the system becomes minimal.
Achieving level four requires datasets that are large enough and have enough validated cases to provide the information needed for the system to “know” the impacts of a production change. The time needed to move from level three to level four varies based on the amount of time it takes to gather the necessary datasets.
Building a Smart Factory
Approaching the move to a Smart Factory as a journey with four stages is critical for one simple reason: there are no shortcuts that can move a manufacturer from level one to level four. Those that have tried find their systems have so much process and data variability that they quickly become mired in complexity. Level three and level four systems require a huge amount of data, which can only be generated and made useful in level two. A step-by-step approach allows manufacturers to progress through a natural evolution. In the earlier levels, they learn more about data systems in general and the data they need for their specific processes. As they learn, they begin to amass the datasets they need to enable the system to identify and execute production-process improvements based on data. With this methodical approach, manufacturers will build a Smart Factory more quickly and with less frustration.
Research services provided by Patricia Panchak.
