Is Jev is a game changer for robotics?
We gave Jev a robot body and handed it complex tasks across navigation, spatial reasoning, and world geometry
120 different real + simulated tasks and environments benchmarking performance against Dimcode, Astra, Fable, Opus, and 5.6
We graded against speed, cost, tokens, # collisions, path quality
Code, Data, and Paper dropping tomorrow. The results were surprising.
Announcing our Gold sponsor challenges.
Dimensional: Best Use of Dimensional
🏆 Prizes
🥇1st place: Unitree Go2 per team + interview opportunity for internship or full-time
🥈2nd place: $500 cash per team + interview opportunity for internship or full-time
Elastic: Find the
Introducing the Dimensional Hardware Grant Program
We’re giving away quadrupeds and arms to devs and creators to build with dimOS
Deploy robot agents as easily as software
Apply below, reply with your project ideas and we’ll fast track your application
As I leave SF after a short stay, I’m more excited to be a builder than ever.
Had the fortune of visiting @fdotinc (thanks @ruslanjabari) and @dimensionalos (this is their current office).
Really humbled by the level of talent here.
This place is 🔥
I'm hiring a head of storytelling to document our mission as we scale to 1000s of shipped robots and build the application layer for atoms
Profound taste, technical knowledge, and end to end creative execution experience required
After AGI only human stories remain
As someone coming from gamedev, something as simple as planning and executing motion on the arm to this day gives me a feeling of wonder.
It’s essentially just an IK solver from which the joint angles are computed and then applied to the motors. Nothing new.
But it moves physically!
(robotics people may now laugh at me haha)
🏆 1st place, Open Source @lenslist
AR as an interface for robot navigation.
Give move-to commands using hands or voice and the robot moves there autonomously.
Works in a variety of environments and conditions. Stock firmware on the @UnitreeRobotics dog. @dimensionalos powers the entire navigation stack with dynamic pathing and LiDAR streaming.
UI and interactions on @specs AR Glasses.
I’m already working on v2.0 with VPS support (@multiset_ai) to reduce drift on longer distances (>5m) and cleaner DimOS integration.
Gotta say, making this is a bit of a rollercoaster as it’s genuinely hard but also I just want it to exist and seeing it work is the most cool and rewarding thing ever.
Repo: github.com/V4C38/spectacl…
started with 2 people and 2 robots a year ago, now 25 people and 50 robots
we’re officially out of space, so join us this friday to celebrate the end of an era and wreck some hardware
Agent autoresearch on robot navigation and control
Using infinite procedural generated scenes to tune trajectory control, and then rollout in mujoco to match real world physics
Shipping open source next week, supports both arm and humanoid trajectory.
Two days ago VECTOR was a homemade mecanum rover with a Jetson Orin Nano and no brain. Today it explored my workshop on its own.
What I did, step by step:
– wrote a dimOS blueprint for it (@dimensionalos open source robot OS), so their cockpit, 3D viewer and nav stack run on it
the fleet is growing past 1000.
we’re giving away 50 robot dogs to developers just because. drop your github, and i’ll send u an application.
get building. repo below.
Every computing shift has produced a new kind of engineer, this one will too. We wrote about who that person is and what we are building for them. x.com/i/article/2087…
Local path planning week. A few cool steps
- wrote inefficienet path planning of my dreams, full depth first search with orientations, autoresearched a very fast algo with same performance (no surprises it's A*)
- recorded locomotion model I/O and ground position on the robot, replayed it in mujoco, used optuna to tune to match the way a real body behaves
- implemented a judge for trajectory (maximum speed without colisions, autoresearched the control algo
The cameras scan your back, YOLO11-pose + YOLOE models segment the scan live into zones: shoulders, upper back, lower back.
Then the two arms with padded end effectors work through, coordinating on a real person.
If a weekend hackathon is enough to get a working demo, imagine what a team could ship in a couple of weeks.
This massage robot was built in 72 hours with dimOS by students who had never worked with a robot before.
An AgileX Nero dual arm with binocular cameras that maps your back and gives you a two handed massage.
2K Followers 3K Following🚀 I turn coaches & speakers into media-verified authorities.
GNAP Media | PR, Press, Google Knowledge Panels.
DM me “FEATURE” to get published.
16 Followers 148 FollowingLeveling up in real estate deals, trading the charts, diving deep into AI innovation, and crushing games when the markets close.
1K Followers 6K FollowingMachine Learning Engineer building in the healthcare and insurance space.
Co-founder of @tinkerhub, a non-profit educational initiative.
2K Followers 4K Following☀️🌊 SoCal born x raised
▶️usmc vet now in tech freelancing +
⏩️BUILDING random & bespoke stuff w/ai +3D
⏺️I post tech/ai, sports, ufo/phenomena
2K Followers 4K Following💻Prototype Architect @awscloud
🌥Cloud, GenAI/AIML enthusiast
🎚Stage lighting passionate
📌Opinions here my own ☄
he / him
Follow me to grow together!