This startup raised $5 million to bring the blockchain to the Internet of things

August 18, 2015, 12:38 PM UTC
Farming alfalfa harvest
Farming alfalfa harvest
Photograph by Thorney Lieberman—Getty Images

Filament is a startup that is taking two of the most overhyped ideas in the tech community—the block chain and the Internet of things—and applying them to the most boring problems the world has ever seen. Gathering data from farms, mines, oil platforms and other remote or highly secure places.

The combination could prove to be a powerful one because monitoring remote assets like oil wells or mining equipment is expensive whether you are using people driving around to manually check gear or trying to use sensitive electronic equipment and a pricey a satellite internet connection.

Instead Filament has built a rugged sensor package that it calls a Tap, and technology network that is the real secret sauce of the operation that allows its sensors to conduct business even when they aren’t actually connected to the internet. The company has attracted an array of investors who have put $5 million into the company, a graduate of the Techstars program. Bullpen Capital led the round with Verizon Ventures, Crosslink Capital, Samsung Ventures, Digital Currency Group, Haystack, Working Lab Capital, Techstars and others participating.

To build its technology, Filament is using a series of protocols that include the blockchain transaction database behind Bitcoin; BitTorrent, the popular peer-to-peer file sharing software; Jose, a contract management protocol that is also used in the OAuth authentication service that lets people use their Facebook ID to log in and manage permissions to other sites around the web; TMesh, a long-range mesh networking technology and Telehash for private messaging.


This cluster of technologies is what enables the Taps to perform some pretty compelling stunts, such as send small amounts of data up to 9 miles between Taps and keep a contract inside a sensor for a year or so even if that sensor isn’t connected to the Internet. In practical terms, that might mean that the sensor in a field gathering soil data might share that data with other sensors in nearby fields belonging to other farmers based on permissions the soil sensor has to share that data. Or it could be something a bit more complicated like a robotic seed tilling machine sensing that it was low on seed and ordering up another bag from inventory based on a “contract” it has with the dispensing system inside a shed on the property.

The potential use cases are hugely varied, and the idea of using a decentralized infrastructure is fairly novel. Both IBM and Samsung have tested out using a variation of the blockchain technology for storing data in decentralized networks for connected devices. The idea is that sending all of that data to the cloud and storing it for a decade or so doesn’t always make economic sense, so why not let the transactions and accounting for them happen on the devices themselves?

That’s where the blockchain and these other protocols come in. The blockchain is a great way to store information about a transaction in a distributed manner, and because its built into the devices there’s no infrastructure to support for years on end. When combined with mesh radio technologies such as TMesh it also becomes a good way to build out a network of devices that can communicate with each other even when they don’t have connectivity.


When designing the use case for the network, the engineer might choose to bring in a satellite radio once a week that can connect to the internet and upload the data from the remote network, or she might engineer the network to keep transferring messages until those messages reach a point that is connected to the internet. This could mean spacing out sensors until they get into range of a cell tower in a remote area for example. The TMesh radio protocol can travel up to 9 miles, but at very low data rates (think 300 bits per second at those distances) so something like a sensor reporting in every hour with a data point would be fine.

Eric Jennings, the CEO of Filament, told Fortune that each sensor costs about $25, but the cost drops quickly based on the volumes ordered. Most clients, which he didn’t name, pay between $10 and $12 per Tap he told me. The funds raised have mostly gone into producing the Taps, which customers should start deploying in large production volumes. He said current customers range from large Fortune 50 to smaller companies in oil and gas, manufacturing, agriculture and mining.

Subscribe to Well Adjusted, our newsletter full of simple strategies to work smarter and live better, from the Fortune Well team. Sign up today.

Read More

Artificial IntelligenceCryptocurrencyMetaverseCybersecurityTech Forward