When Drofika Labs pitched last year at Atlanta Startup Battle, they were selling the crowd on a drone analytics platform that automated flights and provided data to farmers. Though they didn’t win, shortly after, they landed a substantial farm land analysis contract.
But they quickly realized they couldn’t afford to work on the project, due to the cost of storing the large amounts of data they would collect.
“We were a tiny startup with our applications sitting in AWS… we now had an opportunity to work with a great client, but we couldn’t afford to work with them because of the amount of data we were processing through AWS. We’d be spending more money processing the data than getting paid by the client,” the startup’s CEO Kevin Mobolade tells Hypepotamus.
That critical moment when they had to walk away from the client led them to another idea: what has been dubbed “fog computing.”
“There’s so much underutilized computing power all around us. All we wanted to do is leverage that,” says Mobolade. He explains that 50 percent of the cost for data centers goes towards simply the cost of the electricity required to run it.
In a world where fast bandwidth is limited (after all, the U.S. ranks 35th in the world for bandwidth per user) and as smart and connected devices continue to advance, we’ll have more issues storing and accessing data. More companies than ever are relying on data centers, but this isn’t realistic for individual consumers or small startups.
Fog computing allows IoT companies to reap the benefits of the cloud without moving the data off-site. Instead, it keeps the data closer by using local computers and devices already owned by the company. This allows them to process data quicker and more efficiently, without actually storing it.
Dubbed “Airbnb for Computing Power,” Drofika’s blockchain-powered platform and marketplace turns any company’s hardware into small, on-site data centers. In the end, putting underutilized computer power results in lower costs, as well as more control over their own data.
The setup is far simpler, as well. Similar to a virtual machine environment set up, Drofika requires access to upload the application and the IP addresses of the devices that will be used.
One of the devices functions as the master to give Drofika remote access to push out updates and performance optimization.
“By solving this problem, we started understanding the market and realized how many businesses have the same issue of not being able to control their cloud costs,” says Mobolade.
One of their target clients is hospitals, which are starting to require more data processing power thanks to the spread of IoT solutions for administering drugs, smart monitors, sensors and more. This solution helps them get a lower response time in a situation where providers are literally saving lives.
The cost savings is also demonstrable. Mobolade says that one current client spends over $100,000 a month on a training model hosted in AWS that he may or may not use that month. With Drofika, billing only starts when you have an instance of using the data.
“We’re not trying to replace AWS or Azure, but instead offer a third option to try to offload the not-super-critical tasks that people are doing like machine learning, training models, and more.”
The startup is testing a few revenue models including a base fee per device for installation and, similar to AWS, a usage fee. Currently, they have signed on four paid pilots while they’re in beta and will be launching to the public in early 2019.
They’re currently in the midst of raising a seed round that’s mostly committed. “We want to focus on driving revenue and go from there,” says Mobolade.