Transmitting Power Through the Body

The Power-over-Skin ring containing a Bluetooth joystick controller, powered by a small receiver board and electrodes. (Image: Carnegie Mellon University)

People with diabetes rely on continuous glucose monitors to keep track of their blood sugar, but eventually the monitor's batteries need to charge. The same is true for a pacemaker or any mobile device, like a fitness tracker. And batteries are bulky and require regular maintenance.

To free wearable tech from these burdens, researchers at Carnegie Mellon University's School of Computer Science developed Power-over-Skin, which allows electricity to travel through the human body and could one day power battery-free devices from head to toe.

"We can expect all our electronics to keep improving," said Andy Kong, part of the team that developed Power-over-Skin. "New releases, such as smartwatches and glasses, will be able to do so much more, but it will always be difficult to get electronics onto the body because people have to think about charging them. Power-over-Skin opens the door to making these devices invisible, allowing them to do their jobs without you noticing, which is how health monitoring should work."

Still in its early stages, Power-over-Skin allows researchers to design and implement new methods of transmitting power frequencies through the human body. In the study, researchers powered small objects like LED lights, but they envision powering smart glasses or other wearables in the future.

Kong, who earned his bachelor's degree from SCS and returned to the Human-Computer Interaction Institute (HCII) to work on Power-over-Skin in the Future Interfaces Group, said commercially available health-monitoring devices are often placed on the wrist, hand or chest for convenience and to accommodate easy removal. Without a battery, a small health-monitoring device could be embedded into something as unobtrusive as an earring.

Kong worked with HCII Associate Professor Chris Harrison and doctoral student Daehwa Kim to develop Power-over-Skin. Prior work demonstrated that the human body can efficiently transmit 40 MHz radio frequency (RF) without losing too much power to the air. The CMU researchers used a single battery-powered transmitter that's worn on the body to send power to receivers — objects like a Bluetooth joystick embedded into a ring and a light-up earring. While study participants wore these devices in locations ranging from the wrist to the ankle, researchers noted a correlation between the power the devices received and their distance from the transmitter. The closer the transmitter, the more power a receiver got.

Here is an exclusive Tech Briefs interview, edited for length and clarity, with Kong.

Tech Briefs: What was the catalyst for this project?

Kong: Power-over-Skin was inspired by my frustrations with charging multiple devices — I had accumulated earbuds, phone, laptop, and a smartwatch, all of which required daily or weekly maintenance. And while charging is not difficult, I thought that the convenience of the new devices (airpods vs wired earbuds, smartwatch vs casio watch) came at a hidden cost of unreliability (sometimes they're dead, when the older generation of devices could last years).

At the time I was working on wireless power transfer, using RF transmitted through the air. This requires base stations to be installed everywhere you would want to use the power, which I thought was unscalable. Around this time I discovered Prof. Jiamin Li's excellent nature paper using the skin as a medium to transfer a little bit of power, and wanted to see if I could improve it. In this case, all we'd have to do is add a single base station to the body, and all devices I touched would be able to get power off of that. This was how I wanted devices to operate in the future, so I did an initial replication and started improving it.

Tech Briefs: Can you explain in simple terms how Power-over-Skin works please?

Kong: The nearest technology to Power-over-Skin is radio. For instance, when a car receives radio waves, the station is miles away. Using a powerful transmitter and by picking a frequency that travels well through air, it is able to send audio data through the air from their station antenna to your car antenna, which gets boosted and played as audio.

Power-over-Skin is similar, except instead of picking a frequency that travels well through air, we pick one that travels well through the body. The transmitter is powerful and must use a battery, but because the transmitted signal is so strong, the receiver actually doesn't need its own battery, and can instead directly use the power that it sees from contacting the skin.

If this is still confusing, there's a one-minute segment of my talk which is pretty informative:

Tech Briefs: What was the biggest technical challenge you faced while developing Power-over-Skin?

Kong: The biggest challenge was in optimizing and refining the power transmission path. Unlike radios and air, the body is not good at transmitting electrical signals. It heavily attenuates them. Because of this, if the transmitter puts out 100mW, we would be happy to see 100uW on the receiver side. And since there is a recommended limit for how much RF power that a human body can be exposed to, the challenge becomes what signal waveform should we send to maximize power transfer from the transmitter, and what tricks can we use on the receiver to amplify a signal when we have no power.

Because the human body is so hard to model accurately, RF simulation software has limited use, meaning most of the optimization testing had to be done by hand. To try multiple configurations of matching networks, I had to go to the workshop, desolder and resolder a sub-millimeter component, then go back to the testing bench to see the effect. This was very slow.

Another challenge was measuring our power transmission without disturbing the circuit. Usually in electronics, one uses a multimeter or oscilloscope to easily read voltages and debug circuits. However, in our circuit, adding a probe noticeably improved the performance, because adding any conductor increased the power coupling between the receiver and transmitter. Coming up with a new way to measure without affecting the device was a surprise we had to deal with to make real improvements on the circuit.

Tech Briefs: What are your next steps?

Kong: 10x-ing transmitter power would really extend applications beyond simple trackers, but requires a large engineering effort on the RF side of things. Since there are people who are good at this and I am not, this would take an inordinate amount of time so there aren't currently plans to do this.

An effort which I'm more interested in is sending both power and data over the skin. We're currently using Power-over-Skin to transmit sensor data with a Bluetooth antenna — this costs a lot of energy every time we turn on the device. If we could instead send this data back over the skin, we save a lot of energy on communications and could double our energy budget for sensing or for outputs like audio / LEDs. It's also pretty elegant to send everything just over the skin.

We've made initial prototypes for this but need more time to work on decoding this data.