SENSOR measures the signals from Fiber Bragg Grating sensors and other optical sensors embedded within the active chemistry of the battery. (Image courtesy of ThinkStock).
In the pursuit of more efficient and low-cost designs, a growing number of batteries are turning to embedded fiber-optic sensors and machine learning to optimize battery charge. The Palo Alto Research Center (PARC) is partnering with LG Chem Power to develop such a battery management system, with an initial focus on lithium-ion battery packs used in hybrid and electric vehicles.
The Smart Embedded Network of Sensors with Optical Readout (SENSOR) system is capable of monitoring cell degradation and health information, or SOX, as well as predicting remaining battery life. During initial tests, the system demonstrated “2.5% or better SOX accuracy across various xEV use-cases” at both the cell and module levels, said Ajay Raghavan, one of the SENSOR project managers, in a statement.
The SENSOR system is designed with wavelength-shift detection technology, which measures the signals from optical sensors embedded within the active chemistry of the battery. With resolution down to 30 fm and speeds up to the KHz level, this approach yields information about the battery’s state of charge as well as advanced warning of failure. These measurements are transmitted to a read-out unit and evaluated by machine learning algorithms to provide real-time performance management.
PARC recently completed testing at the module level of electric car batteries, following tests on individual lithium-ion cells. The tests on the cellular level were reported to the United States Department of Energy’s Advanced Research Projects Agency-Energy (ARPA-E), which funds the project. The project falls under the Advanced Management and Protection of Energy Storage Devices (AMPED) program.
According to the ARPA-E project abstract, the capabilities of the new system could reduce the form factor of hybrid and electric car batteries by more than 25%, resulting in a lower manufacturing cost at the same energy density.
The next step for the SENSOR system includes additional testing and research into how to scale the technology for larger batteries. Raghavan said that the company will work with original equipment manufacturers to test the technology in hybrid and electric vehicles “as well as to explore its applicability for other energy and structural systems.”
The SENSOR project is only one part of PARC’s Energy Technology Program, which seeks to develop clean and abundant energy for a wide range of technologies. Thus far, it has focused on chemical energy storage for hybrid and electric vehicles, consumer electronics, and the electrical grid; advanced energy conversion devices; wireless sensors; and advanced analytics to maximize energy utilization.