It was September 2013, and Alexey Volochenko had seen what seemed like a big breakthrough in the emerging field of virtual reality: a controller called the Sixense Stem. Volochenko had been an early adopter of the Oculus Rift — he’d gotten one of its first-generation DK1 development kits, long before Facebook decided the fledgling company was worth $2 billion. Where the Rift had put people’s heads in VR, the Stem promised to track their entire body as well. On the very first day, Volochenko says he put down $300, eagerly anticipating getting his new system in July 2014.
When the Sixense Kickstarter opened, the Oculus Rift DK1 had been shipping for roughly six months. While impressive at the time, the DK1 was far more modest than the consumer-ready Rift that would launch in 2016. It had no external tracking and no controller except the Xbox gamepad — which Oculus co-founder Palmer Luckey admitted wasn’t sufficient. So into the vacuum poured small companies with experimental and sometimes bizarre solutions, many of which turned to Kickstarter for funding.
On Kickstarter, you could find the slippery Virtuix Omni treadmill and the very similar Cyberith Virtualizer, which let people run in VR. You could help fund the Delta Six, a detailed replica gun. Companies often promised compatibility between different products; a company called Tactical Haptics planned to build its Reactive Grip controllers on the Stem tracking system. Together, all these projects let people envision a future where you could act out almost anything you wanted in virtual reality, mixing and matching the perfect full-body VR platform.
But today, Volochenko and thousands of other Stem backers are still waiting. Nearly three years after its projected launch, the Stem is perpetually a few months away; Sixense CEO Amir Rubin currently anticipates shipping in June 2017. Rubin says these delays have resulted in a better product, and some backers are still confident. For others, the STEM is a lost cause — a once-groundbreaking product that became a casualty of modern VR’s fast-moving and chaotic early years.
“I had no second doubts about backing their campaign.”
Unlike many VR companies, Sixense wasn’t a brand-new startup. Founded in 2007, it had worked with gaming hardware maker Razer on the Hydra motion controller, which was released in 2011. The Hydra was a flop at release, but it was rediscovered by Rift development kit owners, who realized they could use it to mimic hands in virtual reality. When Razer’s limited stock quickly sold out, Sixense promised to fill the gap with a new and more powerful successor. “[Sixense] was a real company with experience and [a] track record of shipping production units at this point,” Volochenko recalls. “I had no second doubts about backing their campaign and supporting them.”
The Stem was comprised of one base station and up to five small sensor packs, two of which fit into handheld controllers with an array of buttons, analog sticks, and triggers. A weak magnetic field could precisely detect the position of each sensor over a radius of eight feet, nearly three times the Hydra’s range. Early-bird backers could get the cheapest configuration for $149, and a “maxed-out” version started at $1,000. The Stem didn’t raise as much money as the super-successful Virtuix Omni, which netted a total of $1.1 million. But the controller market was wide open, and it ended with around 2,300 supporters and $600,000 in funding, more than double the $250,000 Sixense had asked for.
For a while, things seemed to be going fine. But shortly before the projected launch, Sixense announced that the Stem would be launching later because of hardware changes. While delays are almost inevitable on Kickstarter, the controller was soon pushed to the end of the year, then to 2015 after the product failed FCC regulatory testing. It eventually passed, but Sixense had missed its scheduled window for production, and the Stem was officially bumped to 2016.
By that time, the world of virtual reality had drastically changed. When the Stem was announced, Sixense was one of VR’s biggest names in a deeply fragmented industry; my former colleague Sean Hollister named both it and Oculus “fairly safe bets” at the end of 2013. Three years later, the Oculus Rift had its own motion controllers, and so did the HTC Vive and PlayStation VR, neither of which even existed when Sixense’s Kickstarter launched. People would have to actively choose the Stem over several better-known options, and it still hadn’t shipped.
“We are an R&D team. We are not an engineering team.”
The problem was that while Sixense had spent many years developing hardware, it had never actually released a consumer product by itself — not even the Hydra, which Razer had a large hand in building. “Razer did all the industrial design and mechanical engineering” on the Hydra, says Rubin now. “We are an R&D team. We are not an engineering team.” As a result, Sixense wasn’t experienced in working with manufacturers overseas, handling regulatory testing, or fine-tuning things like the Stem’s plastic casing.
Other companies have certainly overcome similar hurdles, including Oculus, with its infamous manufacturing woes. But the original Rift development kit had surprisingly simple hardware. The Stem is a multi-part wireless tracking system with two gaming controllers, which ended up being far tougher to design than Sixense anticipated. “When you look at the Xbox controller and the Dualshock Controller, the triggers are a hugely complex and challenging mechanical engineering problem,” says Rubin. As of mid-April, the Stem’s triggers were still being tweaked.
Despite all these setbacks, the Stem wasn’t some piece of vaporware rendering. Sixense still regularly takes prototypes on tour at major trade shows like the Game Developers Conference and CES. The company appeared at CES 2015 with an unofficial VR lightsaber demo that won The Verge’s “Best VR” award. The core system was solid — Sixense just hasn’t been able to turn it into something polished enough to mass-produce.
As the controller’s original niche became increasingly crowded, Sixense began expanding its original scope, aiming for businesses and research institutions as well as home users. It recently debuted a piece of design software called MakeVR for the HTC Vive. In addition to slimming down the system, it added new features, adapting the Stem for use on smartphone-based headsets as well as PCs. “The key focus for Sixense is mobile,” Rubin says now. “Mobile VR is where VR will reach scale, reach the mass market.”
The core system works — it just hasn’t scaled
Sixense is, in fact, well-placed to enter the mobile VR world. Mobile headset makers have started toying with a potentially game-changing feature called inside-out tracking, which gives mobile VR users a range of movement that currently requires external cameras or sensors. But in this system, handheld motion controllers would need a direct line of sight to head-mounted cameras instead — which could present a problem when people do something as simple as raise their arms. With the Stem, you could mount a small base station on the headset itself, which would track hands through magnetic fields instead of cameras. It’s not the only way to solve this problem, but it could be a pretty good one.
And for some Kickstarter backers, even Sixense’s original idea has unique advantages over the Rift and Vive’s tracking systems, which need a direct view of controllers to work smoothly. “The killer feature is magnetic tracking that doesn't require line-of-sight,” says backer Chris Jarram. If he’s playing a game like virtual pool (a real and fairly popular pursuit), having one hand even momentarily obscured can ruin a shot. “These problems won't exist with the Stem,” Jarram says. As long as the system ships, he’ll have a place for it.
For others, though, the moment has passed. Swiss Federal Institute of Technology researcher Markus Zank backed the Stem to use with his department’s Rift, but was frustrated when it failed to arrive in time for a student thesis project in 2015. He’s now satisfied with the HTC Vive, and has been seeking a refund since mid-2016. “The VR market was completely different from when we originally backed them,” says Zank. “And it was clear that they would not be able to bring this product to the market.” Volochenko has also requested a refund several times. And what Sixense sees as exciting new capabilities, he sees as feature creep. “I don't see the point anymore even if they ship the unit now,” he says. “Back in 2014 this would've been a great unit to have; now it’s obsolete hardware with no software support.”
“I don't see the point anymore even if they ship the unit now.”
Sixense’s changes and challenges echo those of other companies with crowdfunded VR peripherals, which have spent years trying to deliver on their ideas while staying relevant. Virtuix decided to sell the Omni primarily to arcades instead of home consumers, and as costs climbed, it canceled all international Kickstarter backers’ orders late last year. Tactical Haptics’ original Kickstarter failed, and it’s spent the past years developing new prototypes for Rift and Vive controllers, before recently getting a $2.2 million investment. Cyberith, which Jarram also backed, began selling business-focused hardware and failed to deliver to backers; Jarram and others are now considering a class action lawsuit against the company.
Rubin says that of the roughly 2,100 people waiting for Stem systems, around 150 have asked for refunds. He’s noncommittal about what the exact policy will be, and several people — including Zank and Volochenko — say Sixense hasn’t responded to their requests. In a March interview with Rev VR Podcast host Reverend Kyle, Rubin said that the final call lay with outside investors who have covered most of what he describes as “millions and millions” of dollars in development costs, but he would recommend offering refunds to backers at the point the Stem was ready to ship. “It's out of my control at the moment,” he said.
Sixense may offer refunds to backers once the Stem is ready to ship
Given the Stem’s long history of missed deadlines, it’s possible that its latest one will slip as well. In a mid-April Kickstarter update, Sixense said it was fixing the final problems with the Stem’s plastic housing, in preparation for its impending launch. Responses were overwhelmingly skeptical and negative. “It's like trying to develop a Betamax player when everyone is using Blu-ray,” one backer complained. “In other news, I've had two children since backing this project,” quipped another. Internet comments aren’t a good barometer of overall support. But there’s also less of the forgiving optimism that you can find in earlier updates, when many people just wanted to know what was going on.
Then again, maybe the Stem finally is on track for release. By early May, Rubin expects to have 20 to 25 prototypes available to backers who pledged $1,000 or more. If this goes well, Sixense’s suppliers will deliver a few hundred finished products and gradually scale up from there. He anticipates getting plastics delivered from China by the end of May, and starting to ship Stem units to all backers in early June.
In the long run, the Stem’s greatest legacy might be as a catalyst, not a product. For many people, including me, using it was one of the first times that virtual reality seemed like a place you could walk around and interact with, not just visit and observe. It encouraged developers to think about natural motion long before big headset makers were promoting it. Even as he admits to the Stem’s problems, Rubin proudly recalls the game developers and VR companies that he says were inspired by trying the system. “It's not about arrogance,” he says. “It's about what VR needed.”