Optimizing the supply chain with analytics and IoT

One major output of this analysis is optimizing processes. From plywood to corrugated boxes to paper napkins, Shah and his team run more than 15,000 machine learning models to calculate the optimal production settings based on current business needs.

“The advanced analytics enabled by SAS allows us to find the optimal balance of speed and quality to maximize profitability,” Shah says. “We’re constantly pushing the envelope of what’s possible with analytics.”

Process optimization doesn’t stop there for the manufacturer. With highly complex facilities, complications can arise. Say, for example, step four of a 10-step manufacturing process is disrupted due to wet materials. Machine learning models will recalculate the next-best production process to save products from being downgraded or scrapped altogether.

Elsewhere in facilities, cameras are used on production lines, allowing employees to watch for problems. With SAS, Georgia-Pacific recently began applying computer vision, a form of AI, to these cameras. This technology automatically detects problems, allowing for corrective action to be taken.

Preventing downtime

Pandemic or not, downtime plagues manufacturers. Georgia-Pacific also has a solution to reduce downtime, thanks to SAS.

Using real-time data from 85,000 vibration sensors, Shah and his team can intervene early when their calculations predict an increased likelihood of part failure or electrical problems. By combining this information with historical data on asset performance, they can inform machine operators what to change to optimize the life of the asset.

Automation capabilities of the SAS platform also help reduce downtime. The company receives alerts when something like a bearing, motor or valve is likely to fail, allowing it to bundle those repairs into a planned maintenance event. Repairing machinery in a controlled environment also allows technicians to operate more safely and efficiently.

“Before we began using these tools, our technicians spent a majority of their time looking for assets that were either failing or about to fail. Now they spend that time fixing problems instead of searching for them,” Bakalar says. “It’s a dramatic change in how they work.”

The efficiency gains the manufacturer has seen are impressive. “Our facilities that use these tools have experienced a 30% reduction in unplanned downtime,” Bakalar says.

Additionally, building analytic models using today’s experienced engineers helps to share knowledge with tomorrow’s new hires. “With SAS, we can essentially codify the knowledge that employees would acquire over their careers and use it to get new employees up to speed quickly,” Bakalar says.

Delivering products on time and in full

As a data-driven organization, Georgia-Pacific applies analytics across a wide spectrum of use cases – not just manufacturing. On the commercial side, the company uses analytics to recommend pricing, determine the best day of the week to ship a certain item, and predict customer churn.

One area that particularly excites Tim Berryman, Vice President of Decision Analytics, is the Internet of Things (IoT). Many Georgia-Pacific manufacturing facilities are in remote locations, where high-speed connectivity often isn’t reliable or cost-effective. As such, the company uses SAS for edge computing – processing that takes place at or near the “edge” of a network, where IoT data is generated or collected – which speeds up decisioning.

IoT also plays a role in creating value for our customers. For example, by putting IoT sensors in smart dispensers, Georgia-Pacific can predict stockouts of items like soap and paper towels before they occur, which improves the customer experience and reduces the cost to inspect restrooms to see if they are out of stock.

On-time and in-full (OTIF) delivery is another key focus area for the Decision Analytics group. Supplying some of the world’s biggest retailers with consumer packaged goods brings huge commercial opportunities, but also stiff penalties for late or incomplete deliveries.

Yet again, analytics drives this capability at Georgia-Pacific. To ensure OTIF deliveries, Berryman’s team uses SAS to detect if an order is likely to be delivered late. When this happens, analysts can develop algorithms to quickly find a solution, and automate the flow of insights back to key stakeholders so the company can deliver on its contracts.

“These tools allow us to catch supply chain issues early and calculate the best course of corrective action to ensure deliveries arrive on time and in full,” Berryman says.

“The Viya platform enables us to easily train employees on how to use machine learning models,” Berryman says. “They can quickly crunch billions of records and start making more valuable business decisions.”

“We also have a strategic partnership with AWS, and the combination of SAS with AWS provides us the necessary flexibility and scalability to put analytic capabilities in the hands of users across the organization,” Bakalar adds.

“Previously, these projects would have involved a large data science effort within Georgia-Pacific or handing it off to outside partners,” Berryman says. “Because SAS makes it easy for people to do their own modeling, it’s translated into millions of dollars in value for us.” 

Partnering with the SAS Education team, Georgia-Pacific has trained more than 150 citizen data scientists, a number that is expected to grow as the company continues to integrate analytics into its operations. For Bakalar, this uplift benefits the company and consumers alike.

“For us, ticking up productivity translates into tremendous value,” Bakalar concludes. “And for consumers, whether it’s lumber going through a kiln or paper towels coming off a production line, the manufacturing quality is better and there’s more supply. Ultimately, SAS helps us make better products and get them to market faster.”

Photos courtesy of Georgia-Pacific.