The influence of Artificial Intelligence (AI) in smart manufacturing is growing rapidly. Artificial Intelligence, according to the ARC Advisory Group, applies to any device that perceives its environment and takes actions that maximize its chance of success toward some goal. This includes a vast range of technologies, such as traditional logic and rules-based systems, that enable computers to solve problems in ways that at least superficially resemble thinking.
According to a recent Accenture Artificial intelligence (AI) research report, corporate profits will increase by an average of 38% by 2035 in large part thanks to a more advanced deployment of Artificial Intelligence into financial, IT and manufacturing applications. But at this early stage of AI implementation, is it still not clear how it will be deployed across many possible use cases. Assessments of risk/reward scenarios are being evaluated and many organizations are unsure of how and when to dip their toes into the AI pool.
Benefits of AI in Smart Manufacturing
The benefits of AI can include performance enhancement, cost control, optimization of processes, shortened product cycle development times, and improved efficiency. The value-add of AI also includes 24x7 availability and the capability of machines to learn through experience. In addition, the cost of entry can be very low (depending on the complexity of the application), and savings can be high as a result of very short payback periods. In that respect, it’s worth distinguishing between the learning phase that can require cloud computing and the operational phase that can be much less demanding in terms of computing.
AI also changes the way machine operators perform their jobs and can help capture the knowledge of skilled workers as they transition into retirement. New generations of workers that come into the industrial workforce will start rejecting antiquated process tools and look towards AI as a source of job enrichment notably through robotic process automation for repetitive human actions.
In effect, AI will represent a new way for humans and machines to work together, to learn about predictive tendencies, and to solve complex problems. For example, the challenge today in managing a process that requires tight control of temperatures, pressures and liquid flows is quite complex and prone to error. Many variables need to be factored in to achieve a successful outcome; too many, in fact, for the human brain to resolve on its own. Now, with AI supporting operational decisions, critical factors such as safety, security, efficiency, productivity and even profitability can be optimized. Another example is how AI can help humans for quality inspection providing them with vision analysis and sound analysis.
For industrial environments, two early AI applications of note
Within the scope of discrete and process manufacturing, asset maintenance is one of the industrial processes that is emerging as an early AI application area. More specifically, organizations are beginning to blend the concept of “predictive” maintenance within their more traditional approaches of “preventive” and “break/fix” maintenance. ARC Advisory Group’s definition of predictive maintenance involves applying condition based monitoring techniques to collect and analyze asset data to better understand asset performance and perform appropriate maintenance before impending issues can negatively impact plant performance, availability, or safety.
One common example involves a variable speed drive (VSD) that is connected to a motor. The intelligence within the VSD gathers data regarding any abnormal behaviors in the operation of the motor and then flags the motor for either repair or replacement before any failure occurs. Therefore, rather than waiting for scheduled “preventive” maintenance to occur, maintenance can now be managed on a condition basis. This both lowers cost and increases yield because an asset is only replaced when it actually needs replacing, and any unanticipated downtime is avoided. Similarly, machine learning executed at the edge can help in early identification of power generation turbine blade damage, pump feedwater valve problem, plant motor coupling approaching failure and bearing seal differential pressure problem.
A second area of AI application involves use of a combination of existing systems and new technologies to control the profitability of the plant operation. When profit control principles are superimposed onto process control, a strategy of profitable efficiency emerges. Real-time accounting (RTA), which utilizes a combination of sensor-based data from the process and financial data to calculate the cost and profit points across industrial processes, is the driver for allowing operators to gain access to profitability data. Thus, algorithms can now help operators make the best decision from both a safety and profitability perspective.
Regardless of the application, when entertaining AI, industrial stakeholders should first focus on the main business problem they are choosing to address. Once the problem is analyzed, then technology providers can help to determine whether AI tools can provide a solution that is capable of addressing the problem.
About your Guest Blogger:
Fabrice Jadot first joined Schneider Electric in 1997 focusing on motor control within R&D as part of the variable speed drives activity, which became a joint venture with Toshiba in 2000 named Schneider Toshiba Inverter. In 2012, he joined the corporate company as the Strategy and Innovation Platforms VP dealing with cross-business technology platforms in the domain of digital services, supervisory control and embedded control.
Today, Fabrice is the Chief Technology Officer for Schneider’s Industry business driving automation system architecture, cyber-security and automation digital transformation (Industrial Internet of Things, Industry 4.0, etc.).