Seeing Machines

30/10/2026

In robotics, understanding people accurately matters. But understanding them fast enough to act matters too.

Running on NVIDIA Jetson Thor, Seeing Machines’ Human Mesh Recovery (HMR) achieves 15.8ms median latency from RGB input, while HMR-Lite achieves just 5.5ms.

This combination of accuracy, low latency and compact model size helps make detailed human geometry available while interactions are actually unfolding.

Explore the benchmarks and technology behind our approach to real-time Human Mesh Recovery in our Technical Paper: https://okt.to/8JOpzd

Our new Technical Paper introduces Human Mesh Recovery (HMR), a core capability enabling robots and other Physical AI systems to build a detailed, real-time understanding of people directly on the device they're running on.

27/10/2026

Last week's article from Big Rigs National Road Transport Newspaper highlights the National Heavy Vehicle Regulator’s call for greater action to tackle driver fatigue, following a 16% increase in recorded fatigue-related offences in the last financial year.

📰 https://okt.to/7V63nt

NHVR Chief Operations Officer Paul Salvati reinforces why must remain a priority across the heavy vehicle industry – and why effective fatigue management requires more than compliance alone.

Technology like is one part of the solution, alongside supportive workplace cultures and a shared commitment to putting safety first. The NHVR’s call is an important reminder that tackling fatigue takes a collective effort, with every part of the industry playing a role in keeping drivers and communities safe.

26/10/2026

Effective intervention is more than recognising impairment. It’s about responding in a way that supports the driver while managing the risk.

Early recognition of behavioural change can create an opportunity for the driver to respond and self-correct before stronger intervention becomes necessary.

As explored in Part 3 of our Technical Paper Series on Intoxication, where there is no immediate external hazard, an initial response can inform the driver and provide an opportunity for recovery. How the driver responds then becomes valuable information in itself, with stable recovery supporting de-escalation and incomplete or absent recovery signalling that a stronger response may be required.

By bringing together awareness of the driver and the driving environment, DMS and ADAS can support a more cooperative relationship between driver and vehicle, with intervention escalating as risk increases.

Read Part 3: Turning integrated awareness into appropriate action - https://okt.to/6URTy8

Part 3 of our Technical Paper Series on Intoxication explores how DMS and ADAS can work together to better understand driver impairment.

23/10/2026

For robots to work safely alongside people, knowing that someone is there is only the beginning.

Seeing Machines’ Human Mesh Recovery (HMR) gives robots a detailed 3D understanding of human posture, position, and movement in real time.
That richer understanding can support applications including proximity assessment, motion forecasting, coordinated handover, and human-aware navigation.

It’s about helping Physical AI systems move beyond detecting people to better understanding them.

Read our Technical Paper to learn more: https://okt.to/z0M5Pt

Our new Technical Paper introduces Human Mesh Recovery (HMR), a core capability enabling robots and other Physical AI systems to build a detailed, real-time understanding of people directly on the device they're running on.

20/10/2026

Last week we joined the National Distracted Driving Coalition’s Technology Transforming Transportation event at Insurance Institute for Highway Safety headquarters in Virginia. 🚗

Our demo vehicle, equipped with Generation 3, showed how our aftermarket and OEM solutions accurately detect distraction and fatigue, in line with Euro NCAP and GSR standards.

JT Griffin presented how the technology works, and Nat Todd gave attendees a hands-on demonstration.

Great to share it with regulators and stakeholders as we work to make sure everyone gets home safely.

You can watch JT Griffin speaking on the “AI Inside the Cabin" panel, alongside Aimee Cox, in the afternoon sessions livestream from 31:51: https://okt.to/3xr8lz

19/10/2026

starts tomorrow. We look forward to showing how Human-Centred AI can deliver real-world safety outcomes, alongside global researchers, safety experts and industry leaders.

is proud to support as the Conference Dinner Sponsor where Mike Lenné will address the delegates. You can also see Mike present in the following sessions.

🎤 Pre-conference workshop: Approaches to Supporting Driver Attention - Tue 20 October @ 08:30-10.15
Hosted by the Massachusetts Institute of Technology (MIT), JLR and CARIAD, Inc. (panel moderator), and supported by Touchstone Evaluations, Inc. with speakers from the University of Jyväskylä, Volvo Group, Renault Group and Seeing Machines.

🎤Panel discussion: The assessment of DMS-based systems – what's needed? - Wed 21 October @ 13:30-14:40
Speakers from Euro NCAP, Massachusetts Institute of Technology (MIT), Smart Eye, Seeing Machines and the University of Leeds.

Our research partners at the Institute for Transport Studies at the University of Leeds, QUT (Queensland University of Technology), and AVR3 will also be presenting joint papers that has collaborated on.

🌟Validation of an evidence accumulation model for driver readiness in assisted driving using driver ocular metrics - Wed 21 October @ 11:00-12:30 - Rafael Gonçalves

🌟 Exogenous Disruptions of Visual Attention and Gaze Abnormality - Wed 21 October @ 11:00-12:30 - Huy Phan

🌟 Age-Dependent Changes in Pupillometric Indicators of Cognitive Load During Hands-Off L2 Driving - Wed 21 October @ 15:50-16:50 - Courtney Goodridge

View more details on the conference website: https://okt.to/BhPHZd

16/10/2026

Impairment isn’t always a single moment that can be detected and addressed. It can develop, persist, or worsen over the course of a journey.

In Part 3 of our Technical Paper Series on Intoxication, we explore why understanding both the magnitude of behavioural change and how it evolves over time can help inform a more appropriate safety response.

Repeated lapses in attention, incomplete recovery, or a failure to respond to an initial prompt may indicate that a driver’s ability to self-correct is diminishing. By combining this understanding of driver state with the surrounding driving context, vehicle systems can better determine when a warning may be enough and when stronger protective intervention is needed.

It’s a shift from simply asking “Is the driver impaired?” to understanding how that impairment is changing, what risk it presents right now, and what response is appropriate.

Explore the technical paper - Turning integrated awareness into appropriate action: https://okt.to/80ZsBa

Part 3 of our Technical Paper Series on Intoxication explores how DMS and ADAS can work together to better understand driver impairment.

14/10/2026

It’s National Coach Week in the UK 🇬🇧

From school trips and holidays to sporting events, tours and everyday journeys, coaches help keep people and communities moving. This , we’re celebrating the drivers and operators who make those journeys happen – and the role safety plays in every one of them.

supports coach operators with real-time protection against two of the biggest risks on the road: fatigue and distraction. With in-cab alerts, 24/7 intervention support and actionable safety insights, Guardian helps protect the people behind the wheel and everyone travelling with them.

Because every journey matters – and everyone deserves to get home safely.

13/10/2026

We're proud to be the official Conference Dinner Sponsor at DDI 2026, the 10th International Conference on Driver Distraction and Inattention, 20–22 October. Turning safety research into real-world outcomes takes collaboration across industry, research and policy and DDI is a key forum to connect these communities.

Our Chief Safety Officer, Mike Lenné, will deliver a short address at the Conference Dinner and speak on two of the conference panels – details below. Our research partners at the Institute for Transport Studies at the University of Leeds, QUT (Queensland University of Technology), and AVR3 will also be presenting talks at DDI 2026.

🎤 𝗣𝗿𝗲-𝗰𝗼𝗻𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝘄𝗼𝗿𝗸𝘀𝗵𝗼𝗽: 𝗔𝗽𝗽𝗿𝗼𝗮𝗰𝗵𝗲𝘀 𝘁𝗼 𝗦𝘂𝗽𝗽𝗼𝗿𝘁𝗶𝗻𝗴 𝗗𝗿𝗶𝘃𝗲𝗿 𝗔𝘁𝘁𝗲𝗻𝘁𝗶𝗼𝗻 - 𝗧𝘂𝗲 𝟮𝟬 𝗢𝗰𝘁𝗼𝗯𝗲𝗿 @ 𝟬𝟴:𝟯𝟬-𝟭𝟬.𝟭𝟱
Hosted by Bryan Reimer and Bruce Mehler from Massachusetts Institute of Technology (MIT), Lee Skrypchuk from JLR, and Alex Noble from CARIAD, Inc. (also the panel moderator), and supported by Linda Angell at Touchstone Evaluations, Inc., this session blends expertise from academia, OEMs and our own experience at Seeing Machines to explore how proactive, context-aware technologies can better support driver attention and situational awareness as vehicles and automation evolve.

🎤𝗣𝗮𝗻𝗲𝗹 𝗱𝗶𝘀𝗰𝘂𝘀𝘀𝗶𝗼𝗻: 𝗧𝗵𝗲 𝗮𝘀𝘀𝗲𝘀𝘀𝗺𝗲𝗻𝘁 𝗼𝗳 𝗗𝗠𝗦-𝗯𝗮𝘀𝗲𝗱 𝘀𝘆𝘀𝘁𝗲𝗺𝘀 – 𝘄𝗵𝗮𝘁'𝘀 𝗻𝗲𝗲𝗱𝗲𝗱? - 𝗪𝗲𝗱 𝟮𝟭 𝗢𝗰𝘁𝗼𝗯𝗲𝗿 @ 𝟭𝟯:𝟯𝟬-𝟭𝟰:𝟰𝟬
Mike joins speakers from Euro NCAP, Massachusetts Institute of Technology (MIT), Smart Eye and the University of Leeds to examine how to evaluate DMS-based safety systems, balancing their safety benefits with driver acceptance, long-term behaviour and potential nuisance.

Learn more about DDI 2026 and the conference program: https://okt.to/qiaKfB

08/10/2026

October is National Safe Work Month in Australia – an opportunity for employers to renew their commitment to safe and healthy workplaces.

And for the transport industry, the road is a significant part of the workplace.

Fatigue, distraction and other driving risks need to be considered alongside the broader WHS risks businesses manage every day.

This year’s theme, Safe Work is for Life, is a simple reminder of what’s at the heart of workplace safety: helping people get home safely – today and for the long term.

Find out more and get involved here: https://okt.to/9lJxVz

Safe Work Australia