Discomfort with the Big AI unknown
Discomfort with the Big AI unknown
This section is part of our roundtable report: Innovation in the Business of Manufacturing
Our participants confront the recent advancements in AI technology and discuss both the opportunities and barriers to its widespread adoption.
According to the World Economic Forum, as artificial intelligence advances, “manufacturing is set for a disruption on the scale of the dawn of automation in the 1950s.”
With labour shortages, supply chain disruptions and market pressures compelling manufacturers to shift from an "if it ain't broke, don't fix it" approach to integrating advanced technology on the production floor, there are concerns about the extent of its capabilities and potential for disruption.
Four ways AI can transform manufacturing
Optimize productivity using applications such as predictive maintenance to increase equipment efficiency and effectiveness; self-optimizing machine and process parameters; machine vision for automated inspection to improve product quality; and autonomous mobile robots for in-plant transportation.
Improve operational sustainability by predicting future energy consumption and emissions, analyzing and identifying equipment responsible for excess energy consumption and emissions, and determining optimal process parameters or production sequences.
Enhance flexibility of supply chain management by increasing operational agility and mitigating the impact of external shocks through demand forecasting, network optimization and production planning.
Augment the workforce by supporting employees in their daily work and automating repetitive tasks, enabling workers to focus on higher value-adding activities.
Automation and AI reduce reliance on scarce labour, lower costs, increase profits
A few years ago, Swenco redesigned tool-and-die machinery and automated processes — to the point where the company’s capacity currently exceeds demand. Vice-President Chris Sweeny says operations can now readily scale, and looking ahead, new laser technologies may further reduce reliance on hard-to-find labour.
With growing pressure to produce quality products for lower costs amid labour shortages, manufacturers are increasingly looking to tech to do more for less and to improve efficiencies.
Kevin Sun, COO of Maneva AI, explains why manufacturing tech is crucial. “If I were to ask my son, ‘Do you want to work at my factory lifting 25 kg bags of sugar, 20 times per batch?’ He would say no! And that’s the problem we’re dealing with. Fewer and fewer workers want to do this kind of labour.”
Technology also offers customer service benefits, as Marcel Pantano can attest. MTD Metro Tool and Die weld automotive parts. However, manual inspections are only 80 percent successful and minor welding defects can result in an expensive customer rejection. So, the company invested in a machine vision system that delivers 100 percent inspection success. The impact? Zero rejections, big savings, an even bigger increase in customer satisfaction, and higher sales and margins.
Expand ERP capabilities with AI integration
For most manufacturers, enterprise resource planning (ERP) systems are essential for integrating key functions throughout their operations. But until recently, most of these systems lacked the agility to respond to fast-changing market conditions, could not handle high volumes of data, and could not provide real-time insights.
Now, ERP systems are integrating AI technologies such as machine learning, natural language processing, computer vision, and advanced analytics to enable manufacturers to navigate today's dynamic markets effectively.
AI technologies can power ERP software to analyze vast amounts of data, automate workflows, streamline processes, optimize inventory levels and operational efficiency, and deliver real-time data and insights for fast decision making.
AI is helping manufacturers move toward real-time consumption and root replenishment. For example, AI vision technology can now monitor output per production line. As the product is being manufactured, the system can automatically subtract raw materials from the warehouse and automatically send a purchase order to the supplier.
Or, if there’s a delay on the daily production line, AI can immediately notify the sales department to alert a customer that production has been delayed by a certain amount of time.
For manufacturers with more than one facility, AI can provide managerial insights for multiple locations. For example, Kevin Sun of Maneva AI says artificial intelligence can be used to compare factory performance, even tracking issues such as downtime and recommending fixes.
“Let’s say factory B has more downtime than factory A. The AI system can issue a report indicating that ‘the response of mechanics in factory B to downtime is 35 percent longer than in factory A. Your cost of downtime is $X. Recommendation: hire an additional mechanic.’”
Instead of managers conducting time-consuming analyses, AI will increasingly digitize data and create relevant — and immediate — thinking points.
AI working as a digital twin on the production floor to conduct QC, optimize ops, save costs
Kevin Sun, COO of Maneva AI, which provides artificial intelligence-powered digital line workers to automate key processes, provides a rundown of digital twin capabilities on the production floor.
AI has traditionally been about digitizing things — seeing what's in the real world and putting it into the cloud or worksheets. Humans still did the work.
In manufacturing facilities, tasks performed by humans are based on visual understanding — determining if there are anomalies among products, checking if something is missing, etc.
AI and digitization technologies can now create a virtual replica of a system or facility called a digital twin. AI vision enables computers and systems to analyze visual data and derive meaningful information. For example, by tapping into existing security cameras, AI machine vision can capture images of products in real-time, which the AI system analyzes to identify defects, irregularities, or imperfections.
Identifying tasks that may be improved with AI vision and defining how AI applications will integrate with production lines is now possible. For example, unlike a traditional sensor, a digital twin could identify a piece of steel going through production as being off specification and precisely determine whether it should pass or fail. AI effectively plays the role of a senior quality control tech engineer.
Digital twin technology can also enable manufacturers to model, simulate, analyze and optimize production processes, resources and operations. This can help to identify opportunities for cost savings [18].
Subscription models improve affordability
One manufacturer remarked that AI automation and digitization don’t have to be a huge investment, even for small manufacturing operations, because monthly subscription models are increasingly available.
Manufacturers can now access entire suites of solutions through subscriptions, making advanced technology more affordable and more accessible to implement than purchasing individual products. There are other advantages, such as maintenance and expert support [19].
Subscriptions provide companies with agility and flexibility, enabling them to scale faster while paying only for the software they use and readily adjusting capacity as needed. Moreover, these models utilize a smaller initial investment instead of paying for technology upfront from the capital expenditure budget.
AI-powered automation eliminates repetitive, dangerous tasks
While several manufacturers say their employees worry about automation replacing workers, the reality is quite different. According to the Future of Jobs report by the World Economic Forum, as many as 97 million new roles may be created by 2025 by adopting automation.
The International Federation of Robotics also reports that fewer than 10 percent of jobs can be automated. Instead of replacing jobs, robots eliminate repetitive and dangerous tasks performed by human workers [20], allowing them to transition to more skilled roles.
And for younger workers, automation is particularly appealing. It enhances their job satisfaction and increases retention.
Employee change management is key for advanced tech integration
Several manufacturers commented on the difficulty of changing employee behaviour to optimize the advantages of new technologies like artificial intelligence.
MNP's Hali Van Vliet emphasizes that increasing employee acceptance and responsibility requires an effective change management process. This refers to a systematic approach to transitioning an organization's goals, people, processes, and technologies.
She emphasizes that communication well before introducing new technology is crucial to encourage employees to accept and adapt to change. This includes early and ongoing communication to explain why the technology is needed, how it supports the business plan, how it will work, what's expected of employees, and how this will impact their work.
Change management communication also involves seeking employees’ input. How do they feel about this? Do they have any concerns? How can the company help?
Table of contents
- Introduction
- Overview of our manufacturing industry and key trends
- Biggest challenges — and best remedies
- Finding good employees
- Amping up performance
- Leveraging Gen Zs effectively
- What manufacturers have to say about Canada’s declining productivity
- Discomfort with the Big AI unknown
- Choosing the right tech for the right results