VR Development

University researchers at KU Leuven in Belgium have shown that monkeys can navigate complex virtual environments using a brain-computer interface (BCI) setup, which remarkably involves relatively little user training.

As reported in New Scientist, three rhesus macaques were implanted with Utah array BCI devices containing 96 electrodes in each of three brain regions: the primary motor cortex and the dorsal and ventral premotor cortices.

While the primary motor cortex is involved in voluntary movement, a region of the brain Elon Musk’s Neuralink taps into through its various animal research models and recent human clinical trials, the premotor cortices are thought to be dedicated to planning, organizing, and initiating those movements.

The key innovation isn’t the hardware itself though, as Utah arrays are widely used in research when reading neuronal activity, but rather the method the study goes about decoding that information, and making it actionable in 3D environments.

In the study, which was lead by KU Leuven’s Peter Janssen, the rhesus macaques were initially trained once from a short passive observation phase, and then were given a variety of virtual tasks while wearing 3D shutter glasses and monitor with stereoscopic images. Tasks included moving various objects in a virtual space, including a sphere, a monkey avatar, and even themselves via a first-person perspective.

As noted by New Scientist, many previous human trials involve asking people to actively think of a physical movement, like raising or lowering a finger to move a cursor on a screen, which is then translated to on-screen movement. Janssen believes however the study’s specific placement of the BCI has accessed what could be a more intuitive connection to movement, potentially requiring less training.

“We cannot ask these monkeys, of course, but we just think that it’s a more intuitive way of controlling an a computer, basically,” Janssen tells New Scientist, who notes that current methods can feel as foreign to implant recipients as “trying to move your ears.”

While the study hopes to pave the way for similar results in humans, which could unlock things like controlling electric wheelchairs, Janssen also believes it could allow people with paralysis to intuitively navigate virtual worlds.

“There’s a bit of work necessary to know exactly where to implant a human because a lot of these areas are not very well known in humans, where they are exactly,” Janssen says. “But once we figure that out, it should be possible. It should actually be easier because you can explain to the human what they are supposed to do.”

Lynx has unveiled the Lynx-R2, a significant upgrade over its original R1 mixed reality standalone which aims to capture the enterprise and prosumers market.

The France-based startup considers R2 is a significant step forward, featuring new aspheric pancake lenses from Hypervision which are said to deliver 126° horizontal field-of-view (FOV)—notably larger than R1’s 90°, or Quest 3’s 110° horizontal FOV.

Paired with dual 2.3K LCD displays delivering more than 24 pixels per degree (PPD) at the center, R2 is said to deliver “crisp text and image rendering for industrial and medical use cases.”

While the new standalone headset features the same flip-up design as its predecessor, R2 is powered by Qualcomm’s Snapdragon XR2 Gen 2, offering substantial gains in GPU and AI performance over R1, which was introduced in 2021 with the older Snapdragon XR2 Gen 1.

Other features including 6DOF head tracking, hand-tracking, controller and ring tracking, plus a full-color four-sensor Sony camera array that also includes depth sensing for advanced computer vision.

Originally planned to ship with Android XR, Lynx-R2 is actually set to launch with Lynx OS following Google’s decision to withdraw support. Lynx OS is however based on Android 14, meaning it can sideload APKs in addition to supporting OpenXR 1.1.

Additionally, Lynx says it will release “all the electronic schematics of the headset motherboard and the mechanical design blueprints,” which is said to allow academics
and hobbyists to freely mod the device.

This will also include raw sensor access so developers can enable their own computer vision applications, as well as full offline functionality for sectors such as defense, healthcare, and industry, Lynx says.

“With the R1, we proved that a small, independent team could build a world-class mixed reality device,” said Stan Larroque, founder and CEO of Lynx Mixed Reality. “With the R2, we are proving that an open ecosystem is not just a philosophy, but provides a superior way to approach these devices. We have listened to 3rd party developers and enterprise users. They didn’t just want more pixels; they wanted a wider field of view, faster processing, and total ownership of their sensors. The R2 delivers just that. I believe the Lynx-R2 is a great VR headset, and will provide the best MR experience.”

There’s no official launch date yet. Lynx says R2 will be available for order “starting this summer” via the official Lynx portal as well as authorized enterprise resellers.

In the meantime, we’re still learning about specs, but this is what Lynx has indicated so far:

Lynx-R2 Specs

FluxPose is a 6DOF tracking solution for full-body tracking that seems to be picking up speed on Kickstarter, having now garnered over $2 million in crowdfunding since its initial launch on November 29th.

The News

FluxPose is a full-body tracking system that’s said to deliver occlusion-free positional tracking without the need of externally mounted base stations or sensors. It does this by way of a wearable beacon, which generates magnetic fields, the team explains on the FluxPose Kickstarter campaign.

“It’s completely occlusion-free, incredibly compact, drift-free, and the trackers last up to 24 hours on a single charge, offering high-end performance in the smallest, lightest form factor possible,” the Logrono, Spain-based team says.

And because the beacon is worn on your body, and automatically synchronizes the tracking space with VR headsets without any additional software, it essentially means the tracking volume moves with you as you move (or more likely, dance) in VR.

Weighing in at 85 grams, the trackers are also impressively compact: a Dorito for scale.

At the time of this writing, the cheapest support tier is the ‘Lite Kit’ for €339 (~$394 USD), which comes with three tracking points (straps sold separately). At the higher end is the ‘Pro Kit’ for €689 (~$800 USD), which includes eight tracking points. Notably, those prices do not include taxes or import tariffs.

VR headset mounts provided through the Kickstarter are said to include Quest 2/3/3S/Pro, Pico 4/4 Ultra, Samsung Galaxy XR, HTC Vive Pro/Pro 2/Focus/XR Elite, Bigscreen Beyond 1/2, Valve Index, and Steam Frame. Backers will have the chance to select the exact headset model on a survey after the Kickstarter ends, and again a few months before delivery.

You can find out more over on the FluxPose Kickstarter, which we’ll be following for the campaign’s remaining 58 days, ending on January 28th, 2026. The earliest delivery is expected in August 2026 for early bird supporters, and October 2026 for late comers to the Kickstarter.

My Take

Magnetically-tracked peripherals aren’t anything new in VR; I’ve seen a number of solutions come and go, with the emphasis mostly on go: Razer Hydra, Sixense Stem, Atraxa, Magic Leap 1 controllers—these implementations seem to be good enough in optimal conditions, but not rock solid across the board.

In short, magnetic trackers position themselves in 3D space by measuring the intensity of the magnetic field in various directions, which (as mentioned above) is generated by a beacon. When the trackers’ measurement point is rotated, the distribution of the magnetic field changes across its various axes, allowing for it to be positionally tracked.

And while those magnetically-tracked peripherals listed above don’t suffer from optical occlusion, they can be affected by external magnetic fields, ferromagnetic materials in the tracking volume, and conductive materials near the emitter or sensor. These things typically reduce tracking quality, making them less reliably accurate than optical (Quest 3) or laser-positioned systems (SteamVR base stations).

Granted, I haven’t tried FluxPose yet, although I don’t think those drawbacks are nearly as important in fully-body tracking than they might be in actual motion controllers, which require much higher accuracy. A few millimeter’s discrepancy in your foot’s position really doesn’t matter as much as it might if you were reaching out and trying to grab something with a magnetically-tracked controller.

Provided Road to VR doesn’t get to go hands-on in the coming months, I’ll be keeping my eyes peeled for videos and articles as we move closer to the campaign’s close next month.

Zhenyuan Yang, Vice President of Technology at Pico parent company ByteDance, reportedly revealed plans for Pico’s next XR headset, which is said to sport a self-developed display chip and 4,000 PPI microOLED display.

The News

According to Chinese news outlet Science and Technology Innovation Board Daily (via Nweon), Yang was speaking at ByteDance’s annual scholarship award ceremony when he mentioned specific plans to release a new Pico XR headset in 2026.

The self-developed chip was started in 2022, Yang reportedly revealed on stage, noting the chip is now in mass production. The chip is said to overcome real-time processing bottlenecks in high-resolution, high-frame-rate mixed reality video, with it capable of reducing system latency to about 12 ms while maintaining high-precision image quality.

It’s also said to improve performance in SLAM, motion compensation, and inverse-distortion workloads, which demand high compute efficiency on low-power devices, Science and Technology Innovation Board Daily reports.

Supposedly slated to launch in 2026, the headset will pair this chip with a custom microOLED display which is said to approach 4,000 PPI—slightly higher than that of Apple Vision Pro’s 3,386 PPI.

According to the report, Pico’s microOLED display reaches an average 40 PPD (over 45 at center), and addresses brightness limitations by incorporating microlens (MLA) technology and optical compensation for uniform color and luminance. Additionally, Pico is also developing its own data-capture systems to train advanced eye-tracking, gesture-tracking, and spatial-understanding models.

Yang emphasized that since 2023, ByteDance has shifted Pico’s strategy away from aggressive content and marketing spending toward long-term technological investment, increasing XR R&D rather than retreating from the market.

“In 2023, we decided to reduce our investment in content and marketing, and instead focus more firmly on our technology strategy,” Yang said (machine translated from Chinese). “This was because the hardware experience of our products was not yet mature enough to support large-scale market applications. This adjustment led to some misunderstandings at the time, with many people saying that ByteDance was no longer pursuing this direction. In fact, quite the opposite.”

This follows an initial report from The Information this summer, which alleged Pico was developing a pair of slim and light MR “goggles,” reportedly codenamed ‘Swan’, which are said to weigh just 100 grams.

My Take

More competition is great, although US-based audiences hoping for a new Vision Pro competitor from Pico may be left waiting.

The company’s headsets are typically only available in China, East and Southeast Asia and Europe—but not in North America, and not for the lack of trying either. An additional stumbling block: Pico headsets have typically been priced above Meta’s equivalents, which has limited appeal in Meta-supported regions.

Still, ByteDance, the parent company behind TikTok and Chinese equivalent platform Douyin, has actually overtaken Meta in revenue, putting the parent company in a better position than ever to bolster its XR platform as a premium offering globally.