Mitigate supply chain volatility with innovation

“Risk management is one of the most critical, and frankly overlooked, use cases right now. There are software solutions that can literally say, ‘Here's the supplier ecosystem to which you're exposed.’”

Jonathan Eaton

Principal, Growth Advisory Services
Grant Thornton Advisors LLC

“Third-party supplier risk management is one of the most critical, and frankly overlooked, use cases right now in manufacturing,” Eaton said. “There are software solutions that can literally say, ‘Here's the supplier ecosystem in which you are trading and here is where you have exposure.’ They use hundreds of data sources to build their models. If I'm buying from Supplier A, which is buying from Supplier B, which is buying from Supplier C, all I know is that I'm buying from A. The problem is, if any of those other companies have issues with cyber hygiene, liquidity, litigation, questionable sources of labor — the list is ad nauseum — then there is undoubtedly going to be supply chain risk. So, the question becomes, ‘How can I and should I mitigate the risk?’”

Overall equipment efficiency (OEE)

Some manufacturers can get more insight from their supervisory control and data acquisition (SCADA) and manufacturing execution systems (MES). “Using information obtainable from SCADA and MES systems, you can predict equipment failure and perform preventative maintenance, avoiding failures and increasing overall equipment efficiency,” Eaton said. “Process mining can also identify gaps for efficiency gains.”

“You can look at machine-level data and get to that predictive maintenance story but, maybe more fundamentally, you can stop downtime,” Chandramouli said. “In a perfect world, we'd like to predict when everything's going to fail and get ahead of it. Some companies are just trying to put out fires, and their capacity is limited because their lines keep going down and they can't get ahead of it.”

“OEE is critical because, if you are not running, you're not absorbing your fixed cost and your price per unit is up,” Eaton said. “Anything that gets to all the variables that are involved in conversion and throughput — those are the things that matter the most, because you're in the guts of the business in terms of how the money's made and where the product is produced. So, working from the inside out, you have to get that right, first of all. You can have the greatest technology in the world, but if you've got conversion issues, then you're on your way to being a not-for-profit.”

Demand-supply planning

When manufacturers are growing quickly, innovative technology can help them adapt their demand-supply planning. “A lot of high‑growth mid‑market companies have a large growth curve kicking in,” Eaton said. “You need to understand demand and how you can use algorithms to be more predictive in how you manage inventory. It's easy to forecast something that has a repeatable demand pattern. What gets harder is to plan for something that has sporadic demand or lumpy demand.”

Sustainability optimization

The sustainability landscape is shifting rapidly, with new regulations and consumer demands creating both challenges and opportunities for manufacturers. For manufacturers, there can be a double bottom-line impact when adopting clean tech innovation. It’s not just checking a box for Greenhouse Gas Protocol Scope 3 reporting – you can potentially transform waste streams into revenue streams.

“What's exciting is how hard-tech startups are revolutionizing this space, using AI to optimize entire supply chains for sustainability, identifying opportunities that many manufacturers didn't even know existed,” Chandramouli said. “We're seeing everything from AI-powered emissions tracking platforms to breakthrough carbon capture technologies that can actually turn waste CO2 into valuable manufacturing inputs. These aren't just compliance tools, they're a source of new revenue growth.”

Procurement

“So many manufacturers are trying to be demand-driven,” Eaton said. “If you can do a better job of understanding where your risks are, and do a better job in the source-to-contract process, you can bridge the way you respond to demand from a supply perspective — link that together, and there's a lot of leverage.”

“Procurement has these dense contracts between companies and their suppliers, with terms and conditions,” Chandramouli said. “Being able to extract data points from a contract, and connect a finance system or an MES, you can have shop floor data. You might have something in your contract about quality, and you can get data from your line that shows the reject rate, then tie that back to procurement negotiation. It's connecting departments that don't often talk to each other, and that data can be hard to bridge.”

Quality assurance

In manufacturing, quality always matters. “If you invest a dollar in quality, that can pay back really fast,” Chandramouli said. There are a lot of good use cases now, with accessible and affordable AI applications — like using computer vision cameras for automated inspection. It's not a complex use case to implement, and your customers will see the benefit for a long time.”

Worker safety

Safety is an issue any time a plant is live — but it’s hard to ensure someone is always watching. Technology can help to fill that need. “For example, there’s one startup that's using computer vision and generative AI to make sure that workers are completing processes according to standards and not doing things that might be unsafe,” Chandramouli said. “They can use cameras to detect activities that would require somebody to go coach that worker and retrain them if necessary.”

Staff supplementation

Skilled staffing shortages are still persistent, especially for office skills where manufacturers have to compete with other industries. “This can be one opportunity to invest in AI,” Chandramouli said. “If you can’t get back to your pre-pandemic workforce, you can use AI workflow automation as a viable staff augmentation strategy.”

Commodity pricing

If commodities are a factor in your finances, then pricing can be part of your day. “There is incredible data out there today on commodity pricing,” Eaton said. “There are even software solutions for hedging commodities. Anybody who is buying a commodity can use artificial intelligence that is predictive on where those markets are headed, what the opportunity is, and how that can be used for a hedge.”

Cost to serve

One of the biggest issues for many supply chain participants is the soaring cost to serve. “We see companies struggle with the cost to serve and fail to understand why it goes up.” Eaton said. “It’s because there's customer churn, there are changes in where the product is positioned and where it needs to go, the conversion costs change, and they don't keep track of it all in a proactive way. You even have changes in freight rates and the availability of shipping assets. With the concept of a digital twin, you can use artificial intelligence to be predictive about what your future cost to serve might be, and where there are opportunities to further improve that.”

You can find other use cases where innovation improves your agility to address supply chain volatility. “One way you can look at this is capacity creation — doing more with same or less,” Dinola said. “Whether that's on the shop floor, back office or elsewhere, there are a tremendous number of use cases for that.”

“It's getting people to understand the value of the technology — getting people to think about the use cases and where the technology can be applicable.”

Tony Dinola

Principal, Technology Modernization Services,
Grant Thornton Advisors LLC

Any one of these use cases can create value, but it’s important to align with your future business strategy. “Think about the low-hanging fruit in your supply chain now, but also what's possible in a year, two years, or beyond as you invest in AI as a platform,” Chandramouli said. “It's getting people to understand the value of the technology — getting people to think about the use cases and where the technology can be applicable,” Dinola said. “You need to get in the mode of building around technology,” Chandramouli added. “You might need to find the right partners or vendors — there are innovative and fast-moving startups that need commercial customers or design partners to validate their AI solutions.”

“You might not have the skill sets internally to do all of it yourself,” Eaton said. “Some of it depends on what you’ve already done, and your level of data maturity, access to data about your operation, and how you are executing today.”