If there’s one big problem that plagues all smartwatches and most fitness trackers, it’s battery life. Device designers are faced with two choices: compromise battery life for features, or compromise features for battery life. Watches packed to the brim with energy-sucking sensors — e.g., the Apple Watch and Huawei Watch — necessarily last little more than a day and a half off a charger. Meanwhile, wearables that forgo color displays, GPS, and countless other accouterments stretch that limit to a week and a half, but at the steep cost of functionality. In other words, there’s no best of both worlds.
But John Renaldi thinks his startup Jiobit will change all that with a revolutionary computer chip that will give rise to what he calls “invisibles” — wearables so small and with batteries so long-lasting that you forget you’re wearing them. “We’re embracing and building the next generation of wearables,” he told Digital Trends. “We’re shooting for two to three weeks of active time with a battery half the size of current devices.”
Wearables aren’t mini smartphones
The Chicago-based Jiobit, which has been operating stealthily for the better part of a year, was founded on a tantalizing concept: Hardware that can leverage a combination of machine learning, contextual awareness, and “selective connectivity” to maximize energy efficiency. “Our approach is not only to rely on context, but to progressively figure out behavior with awareness of different actions in terms of power consumption,” Renaldi said. The secret sauce is optimization.
Conventional wisdom, Renaldi explained, has smartwatch makers stuffing the guts of phones inside watch-sized devices and running inefficient software atop them. A smartwatch might use a wireless radio originally destined for a full-sized handset, for example, and that wireless radio might connect to the internet much more frequently than it should — even every 10 seconds or so, in some cases.
Jiobit’s solution, in essence, is to combine hardware tailor-made for tiny devices with cloud-based machine learning. The company’s in-house solution crams a processor, sensors, and a bevy of radios — cellular, Bluetooth, Wi-Fi, and GPS, to name a few — into a compact motherboard that’s “smaller than any comparable [hardware] on the market,” said Renaldi.
Periodically, the chip sends a stream of location and usage data from its sensors to a dedicated Jiobit server, which uses machine learning-powered software to catalog and process it. The end result is a highly contextual method of power management. If the chip is frequently within range of a nearby smartphone, it’ll be more likely connect to it via Bluetooth in the future rather than, say, cellular or Wi-Fi. Alternatively, if it’s often out of range of companion Bluetooth devices, it’ll favor a cell tower instead. “It starts learning behaviors,” Renaldi said. “The server can communicate with the wearable to influence behavior.”
Given the backend connectivity that underlies Jiobit’s technology — the chips can’t really operate without it, Renaldi said — the company is implementing a robust series of fallback mechanisms. One is a “mesh network” that allows the chips to communicate both their location and current status to “a constellation” of smartphone apps and other nearby chips.
“The software knows where the wearable is, and it knows where all of the smartphones are,” said Renaldi. “It’s like a private virtual network that communications location.” Jiobit is also partnering with a phone provider — Renaldi declined to name which one — capable of switching between multiple mobile carriers (e.g., from Sprint to T-Mobile) in more than 120 countries worldwide. “Our solution allows us to intelligently select the best network and coverage,” he said.
This tech could change wearables forever
The sky’s the limit for the technology, in theory, but Jiobit’s focusing its initial efforts on a specific market: wearables to help keep tabs on kids. That motivation came from a very personal experience in Chicago, Renaldi said. While vacationing with family on a road trip, he and his wife lost track of his son for half an hour.
“It was a very traumatic experience,” he said. “Like having a heart attack.” A parent who buys a product packing Jiobit’s tech would “never have to experience that feeling,” Renaldi said. “It gives them a bit more freedom to wonder off and be a kid, and you the parent the peace of mind knowing where they are.”
To cover server costs, Jiobit’s planning to launch a subscription service alongside the wearable. It’s targeted for pre-sale later this year.
In the intervening months, Jiobit’s been testing the chip with “kids across the country,” he said. The technology could eventually become small enough to fit in clothing, Renaldi said, and other unobtrusive forms of “invisibles” like ankle bracelets, necklaces, and jewelry.
“Some people get very lazy with how they build their [chips] and stack their components,” he said. “They use huge … chips. But our components are much more efficient in terms of space.” That freedom could lead to a paradigm shift in wearables: designers, no longer conscripted by battery and size limitations, could create the smart apparel of their mind’s eye.
Jiobit may be relatively young in years, but it’s hardly a fly-by-night operation. Renaldi, a former Motorola executive and the architect behind Moto Maker, Motorola’s online phone personalization tool, founded the firm with two other former Motorola employees, Ed Marabotto and Roger Ady. Jiobit’s notable list of board members include Lior Ron, the product lead for Google Maps; Jim Wicks, former Senior Vice President of Design at Motorola; and Tyler Hall, Vice President at Samsung’s SmartThings. And the company attracted $1.2 million in funding.
“We’re building momentum,” said Renaldi. “We’ve recruited this all-star team to solve these problems. Wearables are an under-shared and overlooked market that we think is ripe for innovation.”