The lab-to-real-world gap isn't a model problem β€” it's a data problem. Excited to see @LeRobotHF and @huggingface open-source one of the most diverse egocentric datasets yet. Thursday Nov 12th.
🀝 Paid partnership
Hub (@hubdotxyz) profile pictureHub
@hubdotxyzNov 10
Robot policies work in the lab. The real world is where they break. To close that gap, we are open-sourcing the most diverse egocentric dataset with the help of @LeRobotHF. Coming to πŸ€— @huggingface on Thursday November 12th.

Nov 10, 2026 Β· 7:02 PM UTC

40
30
169
44,917
RelevantRecentLikes
Bridging the real-world gap starts with better data. Open-source datasets can accelerate the entire robotics ecosystem.
33
Egocentric data at this scale being open-sourced is huge for the robotics community. Can’t wait to explore it.
1
1
7
Absolutely! Excited to see how the robotics community puts this data to use!
3
Closing the sim-to-real (and lab-to-real) gap starts with better data. Big W for open source robotics.
1
8
Absolutely! Better data is the foundation for bridging the gap between simulation, research, and real-world deployment. Excited to see what the community builds with it!
3
Open data like this can accelerate robotics beyond laboratory limitations
1
1
17
Absolutely! Open data is key to taking robotics from labs into the real world.
8
Finally someone saying it out loud β€” it’s not just about bigger models. Diversity in real-world data is what actually moves the needle.
1
1
8
Absolutely! Bigger models need better, more diverse real-world data to truly make a difference.
4
Lab-to-real gap has always been a data problem more than a model problem. Great to see @LeRobotHF and @huggingface pushing this forward
1
1
8
Absolutely! Better real-world data is key to bridging that gap. Excited to see what comes next!
4
Love seeing more high-quality, diverse egocentric datasets becoming public. The community is going to build some wild stuff on top of this.
1
10
Absolutely! Making diverse, high-quality data openly available is a huge step forward. Excited to see the creative ideas and real-world applications the community comes up with!
5
So important framing the lab to real gap as a data problem and open sourcing diverse egocentric data like this is exactly how we close it
1
1
32
Absolutely! Diverse, real-world data is key to bridging that gap and accelerating robotics progress.
10
More diverse first-person data could make a real difference in how robots handle unfamiliar situations. Open-sourcing it is a great step for the whole robotics community.
1
1
7
Absolutely! Diverse real-world data is key to making robots more adaptable. Great step forward!
4
Bridging the lab-to-real-world gap with open egocentric data is a massive win for robotics.
1
1
22
Absolutely! Open, diverse data is key to bringing robotics into the real world.
2
Huge milestone for closing the real world data gap in robotics
1
1
8
Absolutely. Better real-world data means smarter, more capable robots.
7
Exactly. Better real-world data is what turns impressive lab demos into capable, reliable robots. Excited to see what this dataset unlocks for the open-source robotics community! πŸ€–πŸ”₯
1
1
11
Absolutely! Real-world data is the key to turning research into real-world impact. Excited for what’s ahead!
6
Better real world data is essential for building robots that can perform beyond controlled lab environments. Open source datasets like this are a big step forward.
1
1
10
100% lab demos look clean, but real world is messy, unpredictable, full of edge cases. Open data like that is what actually helps robots learn to handle it.
6
Closing the data gap together diverse open datasets will accelerate real world robotics Cant wait.
1
1
24
Absolutely! Open, diverse data will push robotics forward faster. Exciting times ahead!
5
This is the type of contribution that actually accelerates the field. Excited to see what people cook with it.
1
1
9
Appreciate that! Excited to see how the community builds on it and pushes things even further.
5
Calling the lab to real world gap a data problem fits.
1
1
6
Absolutely! Better real-world data is key to bridging that gap.
3
Spot on, RAZA! Lab-to-real-world gap data problem not model LeRobot Hugging Face diverse egocentric dataset closes gap
1
18
Absolutely! Better, more diverse real-world data is key to bridging that gap and advancing robotics.
1
5
Love seeing more diverse datasets becoming open source.
2
13
Absolutely! Open data drives innovation for everyone.
2
Solving the lab-to-real-world gap with diverse open data is such an exciting step for robotics!
1
34
Absolutely! Open, diverse data is key to making robotics more capable in the real world.
6