
Nexar announces AI pioneer @ylecun is joining our Board of Directors as physical AI scales beyond the lab. The era of simulation is ending. Real-world learning is the new constraint. VIsit here for the full press release: bit.ly/45n8Xek

Turning cars into vision-sensors for a connected, collision-free world.

Nexar announces AI pioneer @ylecun is joining our Board of Directors as physical AI scales beyond the lab. The era of simulation is ending. Real-world learning is the new constraint. VIsit here for the full press release: bit.ly/45n8Xek

The most important moment is the one before impact. Any machine moving through the physical world has to recognize a developing conflict early enough for it to matter. Put your own video to the test with BADAS 2.0 at empiricearth.com/products/ba…

The road was the starting point. The next step is helping anything in motion understand what may happen next. Follow @EmpiricEarth to stay up to date with BADAS 2.0.
Cars are only one place where motion becomes risk. BADAS 2.0 can anticipate conflicts across vehicles, forklifts, robots and drones. All from a single video stream. Bring us a scene it wasn’t built for. Test your own video against BADAS 2.0 at bit.ly/4zhVVMc

Our first NETS as Empiric Earth gave us the chance to introduce the new brand in person and show how the work of @getnexar and @NautoInc comes together around a shared goal of making roads safer. Thank you to everyone who stopped by and joined the conversation. Learn more about Empiric Earth's Fleet Intelligence platform by visiting hubs.li/Q04xMKCq0

A map shows the road. It may not show the work zone that moved this morning. Read what @empiricearth CEO @ZMGreenb has to say about why AI needs to keep learning from a changing world.
One thing transportation taught me is that knowing what a road looked like yesterday won’t tell you what’s happening there now. I wrote about what that means for AI and why we’re building @empiricearth. Read “When AI leaves the screen”: hubs.ly/Q04xyrT50

#Nexar is now part of #EmpiricEarth, bringing Nexar and @NautoInc together to turn what happens on real roads into intelligence people can act on. Follow @EmpiricEarth as we begin our next chapter, with news, research, and product updates.

Movement data can show that a vehicle braked. Video shows what was happening around it. A sudden stop looks identical in the numbers whether it was caused by a red light, debris in the road, or a pedestrian stepping off the curb. Without the visual context, all three get logged as the same kind of event, even though only one of them says anything real about risk. For AV and ADAS teams, that context can help separate common driving events from the rare situations worth adding to training and testing before deployment. Take a look at what your training data might be missing: bit.ly/3SVSWZJ

Excited to announce that Atlas has been named a finalist in the AI-Optimized Data Platform category in the 2026 @thecubetv TechForward Awards. Learn more at siliconangle.com/awards/

Nexar and Nauto have signed a definitive agreement to merge. 300M+ miles every month. 10B+ miles of driving history. 50+ countries. Built independently of any single manufacturer. A stronger independent foundation for Physical AI and understanding what actually happens in the real world. bit.ly/3QAzgcV

One week out from ITS America in Detroit, and we are still buzzing. Four days at Huntington Place with the leaders, engineers, and public officials building where transportation goes next, having real conversations about AI, autonomy, and safer streets. The energy never let up. During the conference, our own Mark Pawloski sat down with @Local4News on what Nexar is and how we are helping adress the issues above. Full segment attached below. Watch it, and visit here for a detailed walkthrough with the team and to learn more: nexar-ai.com/ibm #ITSAmerica #RoadSafety

Again, amazing work from Luc, @ronigoldshmidt , Hamish, and Lorenzo. We couldn't be prouder. Here's to the work ahead, and to all the milestones to come. If you want to learn more on BADAS 2.0, visit here and schedule a dedicated walkthrough: badas.nexar.app/


The center of gravity has moved, too. Language models got less of the attention this year. Physical AI got much more. That's the harder problem: systems that have to act in the real world and answer for what happens when they do, not just describe it. A lot of the strongest work was about models that learn how the physical world actually behaves.

The insight that came up most was about data. There's no shortage of talent in this field anymore. Strong models and sharp researchers were everywhere at the conference. What's rare is the real-world driving data to teach those models what the road actually does. That's the whole game now, and it's why the 100M+ miles our network sees every month are becoming more and more important.

At this year's CVPR, our very own Luc Vincent @ronigoldshmidt , Hamish Scott, and Lorenzo Niccolini presented our paper on BADAS 2.0, Nexar's world model, and took home Best Paper at the AUTOPILOT Workshop. It's a real milestone for the team. But beyond the win, they came back with a few things we wanted to share in the thread.

Until today, "safer than a human driver" was a claim nobody could check. Today, @Lockton and Nexar made it a measurement. Two AV systems driving the same boulevard at the same hour cannot be compared to each other. Each one's safety record exists only on its maker's terms. Its own miles. Its own disengagement counts. Neither can be compared to the human driver in the next lane. The new framework measures AV systems against BADAS 2.0, our collision anticipation model trained on 2 million real-world driving events from public roads. This means "Safer than a human driver" stops being a marketing line. It becomes evidence, on the record, that a third party can check. When the claim has to clear an outside bar, the systems built to clear it get better. And when the systems get safer, the roads do too. For the full release visit: bit.ly/3Sgt6Pi