Cummins Joins Partnership to Find Fuel Savings in the Cloud

The partnership, which also includes Peloton Technology, Peterbilt Motors Company, and the National Renewable Energy Laboratory (NREL), is part of the U.S. Department of Energy’s Next-Generation Energy Technologies for Connected and Autonomous On-Road Vehicles program, also known as NEXTCAR.

"We look forward to applying our expertise and working with the other partners," said Ed Hodzen, Director of Advanced Controls Engineering at Cummins. "We can improve our customers' business through real-time optimization of the powertrain utilizing off-board computational resources."

"This enables the use of sophisticated predictive and optimization algorithms that have not been possible to implement on existing on-board powertrain controllers," he added. "The opportunities for improvement of efficiency with additional information and computational resources are significant."

The Purdue-led team will receive $5 million from the DOE over three years for the project, which begins in March, according to a news release issued by the university.

"This project is specific to developing fuel-saving algorithms that depend on connectivity," said Gregory Shaver, a Purdue professor of mechanical engineering who is leading the team. "Trucks will be connected to the cloud, and they will be connected to each other."

The Purdue-led team will pursue three concepts, according to the school’s news release:

• “On-the-fly recalibration” of the engine and transmission to constantly adapt to new conditions;
• Running model-based control algorithms from the cloud as the vehicle is driving;
• And platooning with enhanced capabilities, such as synchronizing transmission shifting between two trucks.

The vehicles will use forward-looking information revealing changes in road, traffic and driving conditions several miles ahead, according to the university's news release. Trucks will be linked to a cloud-based network operations center, providing access to information from crowd-sourced traffic data, road-grade maps and weather services.

"These vehicles will be driven as if every driver had forward-looking information about what's happening a few miles down the road, what the grades are going to be, where the hills are going to be, what the vehicle in front of them is doing," Shaver said. "They are going to be able to react much more quickly, and safely, than a human driver could."

New algorithms will allow for more effective “platooning” of trucks, in which a pair of trucks position themselves to reduce aerodynamic drag, not unlike the peloton in bicycle racing.

"They come together like that because together they reduce the drag on each other," Shaver said. "Through automation, we want to get the trucks closer together than human drivers could safely drive them, and we can do this because the connectivity and algorithms are inherently faster, and more accurate, than humans."

Peloton Technology links pairs of trucks today through a wireless vehicle-to-vehicle communications link between the throttle and braking systems. This allows the trucks to coordinate speeds to maintain a safe, aerodynamic following distance. Optimizing the trucks' powertrains using information about the road ahead could allow for significant improvements.

Peloton Technology's two-truck platooning system results in average fuel savings of 7 percent today at a following distance of 36 feet, based on 4.5 percent fuel savings for the lead truck and 10 percent for the following truck. Enhanced algorithms could boost the average fuel savings to as much as 13 percent.

"Our objective is to tap into fuel savings that can only be attained by managing the powertrain precisely for the road ahead, and for the specific configuration of the trucks in a platoon," Michael Palmer, Peloton's director of research, said in Purdue's news release. "Cloud connectivity provides information about the road ahead, and the trucks exchange data about their estimated mass and powertrain capabilities. This helps us maintain smooth, efficient platooning through grades and rolling hills."

Purdue said NREL will assist with testing of the trucks. Peterbilt will advance its Platooning Development Program. Shaver's team will demonstrate the technology by the end of the three-year project. The commercial cost for the technology is not to exceed $3,000 per vehicle.

Shaver is based at Purdue's Ray W. Herrick Laboratories. Much of the research will be performed in Herrick's Cummins Power Laboratory. Engineers from Cummins will lead the development of algorithms needed to operate the engine and related systems. The project also will involve four Purdue graduate students.

"This is the ideal marriage of the public and private sectors to the mutual benefit of both," said Suresh Garimella, executive vice president for research and partnerships at Purdue. "We anticipate the result will be a commercial and environmental success."