Edward Johns (@Ed_

Edward Johns

426 posts

Edward Johns

@Ed__Johns

Associate Professor and Director of the Robot Learning Lab at Imperial College London.

London, UK

Joined January 2019

Pinned
Edward Johns
@Ed__Johns
Nov 12, 2025
I'm very excited to finally announce one of the most ambitious projects we've worked on — which makes the front cover of Science Robotics today: ☀️ Learning a Thousand Tasks in a Day ⭐️ Everyday tasks — like those below — can now be learned from a single demonstration each...
00:00
109K
Edward Johns
@Ed__Johns
Nov 29, 2022
When we first started playing around with DALL-E over the summer, we realised that this could be intriguing for robotics... And we can now reveal DALL-E-Bot! (1) Convert scene into text, (2) Generate new image, (3) Recreate image with robot. See: tinyurl.com/rfpz6r94. (1/2)
00:00
Edward Johns
@Ed__Johns
May 30, 2023
At every robotics conference I go to, I am guaranteed that a dancing circle of assorted robots will spontaneously emerge at some point… #ICRA2023
00:00
39K
Edward Johns
@Ed__Johns
Jan 28, 2025
Happy to announce that "Instant Policy" was accepted at ICLR 2025! Below is a single uncut video: the robot is learning several novel tasks instantly, after just one demonstration each ... We've achieved in-context learning in robotics! robot-learning.uk/instant-policy (1/7) 🧵👇
00:00
34K
Edward Johns
@Ed__Johns
Nov 21, 2024
This is a single uncut video, showing a robot learning several tasks instantly, after just one demonstration each ... This is possible because we've now been able to achieve in-context learning for everyday robotics tasks, and I'm very excited to announce our latest paper: 🎆
00:00
75K
Edward Johns
@Ed__Johns
Aug 1, 2022
New paper accepted at IROS 2022! DOME enables real-world one-shot imitation learning, where the learned skill can be deployed immediately after the single demonstration. No need for further data collection or training! Led by Eugene Valassakis. Webpage: robot-learning.uk/dome
00:00
Edward Johns
@Ed__Johns
May 21, 2025
A few years ago, humanoids with legs walking around the ICRA exhibition was the new thing. This time, it’s the year of the hands! Tons and tons of humanoid hands! #ICRA2025
00:00
00:00
00:00
00:00
25K
Edward Johns
@Ed__Johns
Oct 23, 2023
So you say that LLMs can't be used for lower-level robot control, because they aren't directly trained on robotics data? Actually, we found that often, they can! "Language Models as Zero-Shot Trajectory Generators". See: tinyurl.com/58aptcur. I'll explain more below ... 🧵
00:00
91K
Edward Johns
@Ed__Johns
Feb 28, 2023
I rather like this video. My students and I often discuss "How can a robot with a single RGB camera know if it's a small object nearby, or a big object far away?". It's about priors, and this video shows me that my human prior favours big aeroplanes over small drones. Yours too??
Massimo
@Rainmaker1973
Feb 13, 2023
Plane crash narrowly avoided x.com/tuidelescriban…
24K
Edward Johns
@Ed__Johns
Apr 10, 2024
Very excited to announce: Keypoint Action Tokens! We found that LLMs can be repurposed as "imitation learning engines" for robots, by representing both observations & actions as 3D keypoints, and feeding into an LLM for in-context learning. See: robot-learning.uk/keypoint-actio… More 👇
00:00
33K
Edward Johns
@Ed__Johns
Feb 18, 2025
I have Post-Doc, PhD, and Research Assistant positions available, for a new project working with me on dexterous robot learning! Come and join me, my team, and our robots, at Imperial College London! 🤖🧑‍💻🇬🇧💂🤖 See robot-learning.uk/dexterous-robo… for more info. Please share/retweet!
00:00
25K
Edward Johns
@Ed__Johns
Nov 29, 2024
Do you want to teach your robot with one demo, then sit back and watch it instantly perform that task? Last week, we released 🎆Instant Policy🎆, and now we are releasing our code and weights... Please try it out and let us know how you get on! Visit: robot-learning.uk/instant-policy.
00:00
12K
Edward Johns
@Ed__Johns
Feb 4, 2025
R+X was accepted at ICRA 2025! Robots can now do in-context imitation learning, just by observing humans going about their daily lives... No more need to *label* and *train* - just *RETRIEVE* and *EXECUTE*! robot-learning.uk/r-plus-x 🧵👇 (1/5)
00:00
10K
Edward Johns
@Ed__Johns
Oct 29, 2024
Could robots collect and self-label their own data? It would be much easier than manually giving endless demos! New at #CoRL2024: 🦾✌️ MILES: Making Imitation Learning Easy with Self-Supervision Our robot learned to use a key from just one demo! robot-learning.uk/miles More 👇
00:00
12K