Battery-Free Bluetooth Sensor Tag Demo

The beginning of January is an exciting time. New resolutions are made for the coming year and new projects are planned. Our year begins with two events that ignite the passion in people of the technology world. Between the Consumer Electronics Show (CES) and National Retail Federation (NRF), innovation spreads from coast to coast. It's a good time to spot new trends, see where products are headed and what will be shaping the technology scene for the months to come.

January also marks Wiliot's second anniversary. Birthdays are a good time to reflect back on milestones that were achieved during the year. When Wiliot was founded our vision was to create a new class of communication devices that would power themselves, be so inexpensive that they could be disposable and produced in quantities that would eventually grow, not to billions, but to trillions. We wanted to bring cloud connectivity to the things that were previously not yet part of the Internet of Things, such as packaging and products like clothing to which you couldn’t attach a radio. Bluetooth was the obvious standard to start with because the infrastructure is everywhere. Making that device using the same techniques pioneered by the RFID industry was theoretically the best way to achieve the size and cost goals. However, this would require that the entire system be shrunk from a printed circuit board crowded with chips, capacitors, a battery, and a crystal, to a single integrated circuit chip glued to an antenna. These would be assembled in vast quantities by machines designed to perform the whole process with minimal intervention, without expensive “clean rooms”. At the start of 2018 we had some ideas on how to do this, but no proof that it was possible.

The list of design problems was even bigger than eliminating the normal components in a conventionally powered Bluetooth beacon. How do you harvest or recycle the wasted signals from the plethora of Wi-Fi, Bluetooth and cellular products that surround us? Those signals are a thousand times weaker than the power from a solar cell. A solar panel would never fit into regular clothing or pill bottles, so harvesting from light or even heat wasn’t an option. How do you run a system where the power comes and goes in intermittent waves, as it does with radio frequency energy? How do you broadcast without a drift between the transmitter and receiver, in the absence of a reliable source of timing e.g a crystal, an essential component of existing radio systems?

The answers to these questions came from an agile process that is familiar to the world of software developers, where new releases are produced constantly, but foreign in the world of silicon chip fabs. A new chip design is traditionally “taped out” and sent to a fab at a much slower pace, often once a year. The Wiliot engineers taped out five prototype chip designs last year. And each time they were bringing up the silicon, the excitement was palpable in the labs as we became aware we were establishing a series of industry firsts: like when we saw the first ARM Cortex-M micro-controller powered by RF signaling like Wi-Fi. A Micro-Controller Unit, providing a flexible framework for the development of applications and instrumenting programming sequences on the chip, running just on a few micro-watts of power; we knew then our chips were able to live, breathe and think. When the third silicon taped out came back from the fab, the question was would our new “baby” talk.

Read the full story here: www.wiliot.com/blog/demo