Software-Defined Batteries Take Charge

One of the biggest challenges of the the Digital Age is simply making it through the day with all devices running. Smartphones and smartwatches go dead at the least-opportune moments, laptop computers run out of juice on long flights, and the range of electric vehicles is severely limited by battery life and the availability of electrical outlets for recharging. "Over the last 15 years, the technology embedded in devices and machines has increased dramatically and the demands have gone up substantially," observes Ranveer Chandra, principal researcher at Microsoft Research.

Although scientists continue to explore battery technology and look for ways to improve performance, particularly for today’s widely used lithium ion batteries, they increasingly encounter technical and practical barriers. So far, "We have run into limitations in density and there is no battery chemistry that delivers a significant leap in battery life, while meeting all the requirements of a single battery," Chandra says. What is more, as designers and engineers focus on building new devices and miniaturizing existing devices—from sensors to wearables—the challenges increase dramatically. "There has long been a need for a different type of battery or a different approach," says Evangelia Skiani, a researcher at Columbia University.

Enter software-defined batteries (SDB). The concept approaches battery design from a fundamentally different perspective. Instead of attempting to build a single battery to power a device or machine, researchers are looking to use different battery chemistries, along with software and application programming interfaces (APIs), to introduce a far more modular, flexible, and efficient system. "By putting hardware and software together, it’s possible to enable many new applications and significantly improve battery life," states Pan Hu, a researcher and doctoral candidate at the University of Massachusetts, Amherst.

Software Takes Charge

Over the last decade, researchers have explored a wide array of technologies that could revolutionize battery life and bolster device performance. These include concepts such as lithium-air and lithium sulfur chemistries and the use of ultracapacitors. Unfortunately, no major scientific breakthroughs have taken place. Researchers have turned to specialized batteries based on existing but evolving battery technologies such as lithium polymer, lithium cobalt oxide, lithium iron phosphate, and lithium manganese oxide designs. However, as Chandra puts it: "There are simply too many trade-offs for mainstream use. They’re fine for certain tasks and situations and some can charge faster, but they also don’t deliver the same battery life, or they are too bulky or expensive."

Rather than wait for engineers, chemists, and other scientists to develop the perfect battery, researchers have embarked on a different path: mixing and matching batteries with entirely different designs and chemistries to address the specific needs of a device or user. Within an SDB model, the software manages and optimizes the different batteries while combining human input and machine learning to dynamically adapt the underlying algorithm. In this way, it is possible to use several different batteries together and take advantage of the best characteristics of each; the operating system views the separate batteries as a single entity and delivers performance gains.

SDBs could open the door to new approaches and scenarios: bendable batteries that rely on a hybrid approach to power wearable devices far beyond today’s limitations; systems that support high-power workloads by combining features such as a fast charging battery with a high energy density battery–or a "turbo" mode that better matches the needs of the CPU; and two-in-one capabilities that manage separate but connected batteries or devices. For example, some computing devices have an external battery pack located under the keyboard, which supplements the main battery; an SDB system can efficiently decide how much power to draw from each battery, thereby increasing the battery life of the device.

Left to Our Devices

Early research conducted by Chandra and his team demonstrates it is possible to extend the battery life of a wearable device by 250% with an SDB. In some cases, up to a 22% improvement in overall battery life is possible for two-for-one devices. Yet Chandra says the benefits extend far beyond these use cases. Better matching of batteries with performance requirements makes it possible to build more efficient electric automobiles, drones, even computer operating systems. For instance, a calendar app may be able to predict what a person will be doing at a certain moment—such as using videoconferencing—and better manage battery resources for that event.

Microsoft has developed prototypes and showcased the feasibility of the technology. It hopes to plug SDB into a wide range of products and devices in the not-too-distant future. Says Chandra, "The technology opens up a new area of research and opportunities. Designers and engineers are no longer constrained by the chemistry of the base battery. By combining chemistries, it’s possible to innovate and produce a system that’s more dynamic and more configurable. It addresses a problem that has been lingering for ages."

Samuel Greengard is an author and journalist based in West Linn, OR.