←
Applied Intuition1,148 posts
Applied Intuition avatar✓

Applied Intuition ✓

@AppliedInt

Applied Intuition is the physical AI company bringing intelligence to every moving machine on the planet.

◷ Joined August 2017▧ 525 Photos and videos
Applied Intuition avatar
Applied Intuition@AppliedInt·Nov 12

Our CEO and co-founder, @qasar recently shared his takes on how to build a successful company.

0:35
Applied Intuition avatar
Applied Intuition@AppliedInt·Nov 12

As cars become more and more intelligent, the cabin could one day adapt to every driver. So how do you keep a Jeep feeling like a Jeep? @Stellantis’ Chief Design Officer Ralph Gilles and our Head of Design and Marketing Karl Heiselman talk brand identity in the era of AI. appliedintuition.com/blog/st…

1:04
Applied Intuition avatar
Applied Intuition@AppliedInt·Nov 3

End-to-end driving doesn’t need human demonstrations. At millions of steps per second, self-play produces the collisions, near-misses and recoveries that no driving log contains. TerraTransfer++, the latest in our Terra Series, distills that experience into a camera-only planner that outperforms every published method on NAVSIM navtest and navhard: ➡️ 59.4 EPDMS on NAVSIM navhard, ahead of DriveZero-Scale (57.1) ➡️ 95.5 PDMS on NAVSIM navtest ➡️ 0.548 closed-loop HD-Score on HUGSim Paper: arxiv.org/abs/2610.30319 Project: terra-applied.github.io/Terr…

Encord avatar
⟳ Applied Intuition retweeted
Encord@encord_team·Nov 2

“The fantasy that these machines that exist around us are all going to be replaced by new machines. This just doesn’t exist.” - @qasar, CEO of @AppliedInt Qasar’s argument is that the real work in physical AI is retrofitting the billion machines already on the planet. A consumer car runs for 15 years, construction and mining equipment for 30, and defense hardware for six decades, so most of the world’s machines will get their intelligence added, not built in.

1:05
Jay Kapoor avatar
⟳ Applied Intuition retweeted
Jay Kapoor@JayKapoor·Oct 27

I know it feels like a really competitive market, but Physical AI is not a zero-sum game. @AppliedInt CTO Peter Ludwig looked out at 200 founders at Dirty Jobs this year and said every one of them can win, because the markets we build in are really THAT big. He explained why we're seeing an explosion in interest across autonomy "Reinforcement Learning only got fast enough to matter in the last 12 to 18 months" The time is now. EP 104 of CLIMB 👇

22:41
Wei ZHAN avatar
⟳ Applied Intuition retweeted
Wei ZHAN@Wei_ZHAN_·Oct 27

Can we achieve state-of-the-art for mainstream E2E and vectorized driving benchmarks without using any human demonstration data? Terra series at @AppliedInt now tops the NAVSIM, HUGSim, nuPlan val14 and InterPlan benchmarks, powered by self-play RL without imitation, scaling in closed loop with all agents being reactive. 🚘 With TerraZero released a few months ago, we set the SOTA on vectorized benchmarks like nuPlan val14 and InterPlan, powered by our super high-throughput self-play RL framework experiencing 25 centuries of challenging driving scenarios. 📽️ In June this year, TerraTransfer was released as the first of its kind to bring the SOTA vectorized policy from self play (TerraZero) to an E2E policy via image-to-vector encoder alignment, which topped the HUGSim closed-loop score. 🔚 This week, we are thrilled to release TerraTransfer++, a demonstration-free E2E planner setting the SOTA for NAVSIM and HUGSim, by adding planning capability from self-play rollout experience to TerraTransfer. Setting SOTA for public benchmarks is just a beginning, as we have seen the journey from pioneering work like UniAD to successful mass production of E2E ADAS in the industry, but what makes me quite excited is to see the strong signal from a new paradigm with potential to refactorize how the next generation autonomy can be developed, and bring the scalability, safety and performance to the next level. We are the believer of data and scaling, but not just human demonstration data and open-loop scaling like imitation or open-loop RL (AV2.0) - we have the bet on closed-loop scaling with both synthetic and real-world data (AV3.0), and the Terra series is the starting point of the journey towards real product with our bet. Great work from the team @caffe_tt, Weixin Li, Zhouchonghao Wu, @AkshayRangesh, Saarth Bonde, Grantland Hall, @ChenTangMark, Yihan Hu, for creating TerraTransfer and TerraTransfer++! TerraTransfer++ report:  arxiv.org/abs/2610.30319 TerraTransfer++ website: terra-applied.github.io/Terr… TerraTransfer report: arxiv.org/abs/2606.17386

1:17
Applied Intuition avatar
Applied Intuition@AppliedInt·Oct 27

Then we teach it to see. Adapters on a frozen DINOv3 backbone align camera features to the frozen teacher’s latent space. Every target traces back to self-play, so training needs only paired (image, scene-state) frames. No expert trajectories, no online rendering.

Applied Intuition avatar
Applied Intuition@AppliedInt·Oct 27

The key move: decouple learning to drive and plan from learning to see. Self-play in a vectorized sim runs at millions of steps/s and naturally covers collisions, near-misses and recoveries. We distill those rollouts into a flow-matching trajectory planner.

Applied Intuition avatar
Applied Intuition@AppliedInt·Oct 27

E2E driving is bottlenecked by data. Imitation needs fleet-scale logs that barely cover the safety-critical long tail. Closed-loop RL in photoreal sim pays per-step rendering and backbone cost. TerraTransfer++ removes human demonstrations from every stage and achieves state-of-the-art driving performance across open-loop and closed-loop benchmarks. 🧵 Paper: arxiv.org/abs/2610.30319 Website: terra-applied.github.io/Terr…

1:17