After 8 months of building in stealth and testing our infrastructure on 10000 hours of real-world data and hundreds of unique environments, we're bringing @fpv_labs into the open today. FPV Labs started with the following bet - if human data proves to be the underlying factor that determines scaling laws in general-purpose robotics, it will trigger the largest economic transformation in human history, and the underlying infrastructure that captures that data will determine how fast we get there. We will achieve this by building the full-stack infrastructure for capturing, processing, transferring, and evaluating human experience into spatial, temporal, and semantic knowledge for machines. Despite all the research novelty behind ChatGPT, its success can be attributed to one foundational fact - the scaling law of transformers. We believe the same dynamics have made their way into robotics. Recent studies showed task completion rates jumping from 30% to 70% when human demonstration data scaled from 1,000 to 20,000 hours, a log-linear trend that mirrors exactly what we saw in language and vision. Seeing these emergent signs of scaling law curves in robotics, we believe we are entering the era of general-purpose robotics policies, which makes the next few years the most exciting time in the history of this field. But the library of physical interactions required to train general-purpose robot policies does not exist yet. Over the last 8 months, we've seen dozens of companies emerge in this space. We were really happy to see new companies pushing this space forward, but we also saw the same pattern repeat: every egocentric data company was making some tradeoffs between quality, scale, and diversity. We have built FPV labs on the core principle that high-quality data is orders of magnitude more valuable than sheer volume. Case in point, self-driving cars collect thousands of hours of data per day, but only a small fraction of that data is actually useful for training better models. Several studies, like RT-2, have shown that as little as 1% of data improves as much as 25% on task success. The quality and diversity of data matter a lot more than scale, so there is clearly a power law curve in the downstream impact of data. We've spent months obsessing over data quality by building our stack, discarding it, rebuilding it, and iterating until we found a formula that doesn't compromise downstream quality at scale. We believe the downstream impact here is far more profound than most people realize. Workers globally are paid around $60 trillion per year in aggregate, and a lion's share of that compensation goes to physical labor - tasks that require navigating real spaces, manipulating real objects, and negotiating the infinite variability of the physical world. Human-to-robot transfer will be one of the most important infrastructures that will shape our society in the near future, and if it works, the economic impact will dwarf every technology transition that came before it in an exponential manner and lead to the creation of goods and services we can’t imagine today. Our mission is to lay the groundwork for us to transition into this future - the future of abundance. We are deeply grateful to our earliest believers, @paraschopra and @lossfunk, who played a critical role in shaping our thinking.

We benchmarked an iPhone Pro running ARKit against a 24-camera Vicon optical motion-capture system (sub-millimeter accuracy at 200 Hz) to find out if iPhone tracking accuracy is true or a myth. Go read the full essay and let us know what you think.

Can we build generalist robots with zero teleoperation? Come participate in the discussion and weigh in at our ICRA'26 workshop, BeyondTeleop, starting at 8.45 am CEST today (June 5th)! 📍 Strauss 3

Can a commodity smartphone replace a $10,000 data rig?
 We're presenting MobileEgo Anywhere at ICRA 2026 - open infrastructure for collecting hour-long egocentric trajectories on commodity hardware for downstream robot policies. 
 📍 Strauss 3 | 🕒 3:00 - 4:00 PM today

Come say hi 👋

As much as I like ai for life sciences, I also like to stay updated on whats happening at the physical AI space. I have been seeing a lot of startups popping up in this space recently. Most of the startups are working as a data provider / vendor. Essentially a problem of ops and scale. While thats all cool, I really like this open source project from @fpv_labs. I was going through their data, and their visualizer is so cool. I was able to see so many types of modalities in real time, and everything on web. Super cool project. I saw them recently open sourcing around 1 TB data at huggingface. Something worth checking out!! Great work.

I don’t think people are realising how differentiated our recent release is for physical ai - 1. Collected on commodity hardware 2 High fidelity multimodal data with rgb, dense depth maps, imu, room mesh, hand tracking, 6dof poses and dense action labels 3. Fully reproducible by anyone through Stera SDK. The data engine for egocentric data is now open source. huggingface.co/datasets/fpvl…

Been digging through STERA-10M all morning. some notes: 10M FPV samples for training world models, just released by @fpv_labs on Hugging Face. a few reasons I think this matters for embodied AI / spatial intelligence 🧵 huggingface.co/datasets/fpvl…

Democratised infra for high-quality multimodal data -> more data labs -> more data diversity without trade off in downstream quality -> accelerated progress towards embodied foundation models -> accelerated future towards the future of abundance.

Introducing Project Stera by FPV Labs, an open data infra for embodied AI research. Project Stera includes Stera-10M, with 10M frames of long-horizon data with persistent state tracking, and an open-source pipeline that converts raw data into training-ready formats.

LiteFold just got featured in @YourStoryCo It’s been 8 months since Cory and I officially started @try_litefold. We began by building better interfaces and workflows for computational biology, followed by Rosalind, our AI Co-Scientist for therapeutic design. Now we’re gradually moving toward developing our own in-house scientific and AI research focused on next-generation therapeutic models and research environments and partnerships. Still very early. Really cant wait to share whats coming. Story link in the comments

Gave a talk at Microsoft Research on how @lossfunk operates. We're still iterating but so far our style of operating has resulted in four papers accepted in ICML, RLC, ACL and ICLR main conferences ( many workshops). All undergrads, btw! 🚀 Slides below 👇

We don't even design binders!! But we @BioMandrake just won @adaptyvbio × @gembioworkshop's RBX1 design competition - 1 Strong binder out of 322 tested, selected from 12,000 submissions. Couldn't make it to @iclr_conf in Rio. Here's how we did it 👇 open.substack.com/pub/mandra…

Human to robot transfer involves 2 problems that are tightly coupled - capturing how humans interact with the physical world and translating that knowledge into actions a robot can execute. Both depend on one thing - knowing precisely where the camera was in 3D space at every frame.

Lie algebra is a parameterization choice, particularly for non-linear optimization. The blog post itself just discusses simple SE(3) matrices in the setup, but the optimization of the underlying cost function for example - bundle adjustment - usually involves non-linear optimization in different parameter space - the most common choice is the lie algebra space

SLAM / VIO in egocentric settings is pretty hard. Low-texture scenes and rapid head movements break things.