Watch Now


Cummins relies on predictive analytics to navigate pandemic

Planning skill helped guide inventory where it was needed most

As the new executive director of global supply chain planning at Cummins Inc. (NYSE: CMI) Heather Kettelhohn spent just four days in her office. Then the coronavirus began shutting down operations globally.

The Dow Inc. (NYSE: DOW) veteran joined the engine maker in March. She moved from Michigan to Indiana. Working remotely from her living room became the new normal.

FreightWaves Detroit Bureau Chief Alan Adler speaks with Cummins Inc. Executive Director of Global Supply Chain Planning during the FreightWaves North American Supply Chain Summit.

“It was a strange time to join a new company and a new industry,” Kettelhohn said in a fireside chat during the FreightWaves North American Supply Chain Summit. “I really have no idea what is normal for this industry. My entire tenure here has been colored by the pandemic.


“In my old role, I would have known exactly what levers to pull, where to go get data, what we needed to do.”

Dramatic swings

Cummins recorded the biggest quarter-over-quarter drop in demand in its 101-year history in the second quarter.

“We were just kind of aligning to that, figuring out how to preserve cash, how to adapt in this low-demand environment. And then, in the third quarter, we had our biggest quarter-over-quarter historical increase in demand,” she said. “As I look back, probably the biggest challenge is how do we create enough agility to respond to that type of change.”

Cummins had neither the logistics flows nor the planning systems to deal with what amounted to a cold shutdown of its plants in the latter half of March and April. That was followed by maximum overtime in the summer months.


Heather Kettelhohn, executive director of global supply chain planning for Cummins Inc.

“We had to get very creative on the fly,” Kettelhohn said.

The deployment of some visibility and supplier collaboration lessons in 2019 created a broader view of inventory.

“We were able to, if plants needed critical parts, see where those parts were and bring them back for critical engine builds,” she said. “At the same time, we were able to look at inventory parts in a plant and say, ‘It’s actually more critical that we service customers over here’ and move those parts around.”

Predictive analytics

The business impacts of COVID are less severe now because of what Cummins learned. Manufacturing lines were redesigned to create social distance between workers operating in close proximity. Collaborative robots now perform certain tasks without endangering nearby workers.

The long-term solution, Kettelhohn said, is predictive analytics. Forecasting demand by using data collected from telematics and sensors placed on Cummins’ engines helps discern what parts are needed when and where.

“Is there something we can glean from that to predict one month, two months, three months ahead of time [to] make sure we have the right parts in the right locations?”

Prediction is daunting . Cummins sells tens of thousands of parts, has more than 600 wholly owned distributors and 7,600 dealer locations globally.

All of Cummins’ major truck making customers have their own telematics and sensor data. It is not shared with Cummins today.


“I’d love for us at some point to really have an end-to-end from supplier to end customer supply chain,” Kettelhohn said. “I think the whole industry is moving toward how do we [get there]?”

Cummins’ Q3 earnings: A classic V-shaped recovery

Cummins sees $400M in revenue from making hydrogen in 2025

Navistar, Cummins developing fuel cell truck for Werner testing

Click for more FreightWaves articles by Alan Adler.

Alan Adler

Alan Adler is an award-winning journalist who worked for The Associated Press and the Detroit Free Press. He also spent two decades in domestic and international media relations and executive communications with General Motors.