A Four-Part Plan for Smart Manufacturing
ISE Magazine – Volume: 49, Number: 07
By Andrew Kusiak
Manufacturing has been evolving and becoming more sophisticated, automated and computerized since its inception. Smart manufacturing is an emerging form of production that integrates the manufacturing assets of today and tomorrow with sensors, computing platforms, communication technology and data intensive modeling, control, simulation and predictive engineering. Smart manufacturing uses the concepts of cyber-physical systems, the internet of things, service-oriented computing, cloud-computing, artificial intelligence, and data science. Once implemented, these technologies would make smart manufacturing the hallmark of the next industrial revolution.
The digital economy promises to revolutionize manufacturing. Increasing volumes of data and information are being collected on materials, products and equipment. Data analytics and predictive computer models are being developed to anticipate the failure of a mechanical component or disruption of a supply chain. Tracing product failures back to the source of error or faulty components enables problems to be fixed swiftly, avoiding expensive recalls and litigation. Quick improvements save resources and energy for the enterprises involved.
The shape of all industries will change in a decade and beyond. Some forms of manufacturing will be distributed, others centralized. Instead of producing a restricted range of items and shipping them around the world, many products will be customized and manufactured locally. Personalized drugs or implants could be 3-D printed at the hospital rather than produced at remote locations.
Local vs. global
A polarization of the coupling between manufacturing assets and the enterprise may take place. For corporations with generic manufacturing processes, the coupling will become weaker. Businesses versed in product and process innovation will see a strong coupling. The weakly coupled manufacturing assets may follow the path of information technology and other services that get outsourced.
But making manufacturing “smart” is challenging. Enterprises and supply chains operate globally, while manufacturing is optimized locally. This largely is due to the computer and manufacturing technology lacking the necessary connectivity and the organization structure, which is designed to serve a local enterprise. No single corporation can change complex, interdependent systems based on markets and emerging technologies that are uncertain.
To boost progress four things are required. A platform is needed for publishing industrial problems and solutions. Physical spaces and cyber-laboratories should be provided to enhance such collaborations. Researchers and industry need to share data to develop models about manufacturing. Smart manufacturing friendly policies are needed. But before delving into the four-part plan, let’s look at the current state, along with how the landscape is changing.
Sensors and wireless technologies deployed throughout manufacturing processes will enable collection of a wide range of data, from materials and process parameters, including material composition and properties through the health status of manufacturing equipment reflected in the temperature and vibration pattern to the anticipated product quality and information about products, customers and suppliers. Right now, the collection of such data varies quite a bit across industries. While manufacturers have long monitored productivity, processes and product quality, the upcoming possibilities are much greater.
Computer modeling of processes at various levels of an enterprise and integration of data from diverse sources would bring insights into, for example, risks that critical components might not be delivered on time because of severe weather conditions and issues with manufacturing quality. Interactions between phenomena in disparate domains such as materials, processes, productivity and product quality could be explored. New technologies create even more challenging problems of predicting the quality and properties of a printed part given the variability of material, geometry and the process itself.
Most current analysis performed in industry is based on the data collected for purposes other than modeling, ranging from process control to meeting accounting and regulatory needs. Creative activities and decision making will increasingly take place in a participatory environment, which will enhance innovation and lead to better decisions. Smart manufacturers will share access to manufacturing hardware, while the technology details and know-how of the manufacturing systems will be protected. Buying manufacturing capacity rather than subcontracting components may offers greater flexibility in production management. With smart manufacturing, the service and contract models likely will be deployed at larger scales and with more sophistication.
Smart manufacturing should enhance sustainability by focusing on materials, manufacturing processes, energy and pollutants. Organic and biomaterials may be produced using little energy in an environmentally friendly way.
Minimizing transportation distances for products and components will reduce costs and environmental harms. There are two broad industrial transport categories: internal, which involves moving items using specialized factory equipment or various trucks, and external, which moves the materials, components and products across the supply chain on ground, water and air transport.
Changing global manufacturing supply, value and profit chains requires collaboration across industrial sectors. International collaboration efforts have been initiated in the past. Perhaps the main lesson learned from these collaborative efforts was that besides technology, trust, will, conviction and policies are needed to accomplish a common good.
The expansion of globalization has eased some of the reservations that industry might have had toward collaboration. This, in concert with developments in computer, communication and manufacturing technology as well as advances in artificial intelligence and data science, have served as disruptors for transforming the previously initiated intelligent manufacturing efforts into smart manufacturing.
The four-part plan
The progress in smart manufacturing would benefit from four actions:
1. Establish problem definition networks.
2. Develop cyber-platforms of modeling and innovation.
3. Make data sharing a reality.
4. Enact policies that are friendly to smart manufacturing.
By acting as one, industry, government and academia can make the next industrial revolution a reality.