XPeng’s IRON Just Got a HUGE Human-Like Upgrade

Infinity Robotic
4 followers
AI Generated·May contain errors
Imagine walking into a car showroom and meeting a humanoid robot for the first time.
You tell it a little about yourself and which car you are interested in.
Later, you return wearing completely different clothes. Yet, the robot still recognizes you
and remembers what you talked about during your first visit.
That is what Xpeng is now trying to build with its iron humanoid robot.
Its latest experiment moves beyond the walking and dancing demonstrations we normally see
and focuses on something much harder. Xpeng wants iron to understand who is around it
and remember previous interactions while responding naturally as conversations change.
Later, we are going from a robot learning how to understand people to another humanoid
that spent the equivalent of 140 years learning how to play football against them.
Hey guys, Alfie here. Welcome back to the AI Nexus.
The footage was released from Xpeng's robotics lab on September 22nd.
CEO He Xiaopeng described the progress as moving from human-like motion toward human-like interaction.
Xpeng placed three people around iron and gave each one a different customer profile.
One wanted a family SUV for camping. Another preferred the GX.
The third was a new driver looking for something attractive and easy to park.
Those introductions give iron information it can use later.
But things become more interesting when everyone starts moving around the room.
Iron continues facing whoever is speaking and changes the way it turns depending on
where that person is standing. This behavior comes from several
technologies working together rather than one simple tracking feature.
Nine microphones first help iron locate the direction of a voice.
Visual information such as lip movement then helps confirm who is actually speaking.
Once the AI identifies that person, its motion strategy decides how much of the robot should move.
A small change may only require the head, while a wider angle brings in the waist.
For a much larger change, iron repositions its entire body.
Humans make these adjustments without thinking.
But iron has to combine sound localization and visual perception with physical motion control.
This means the conversation is connected directly to where the robot places its attention
instead of working like a chatbot that simply happens to have a humanoid body.
The three people eventually swap positions and change clothes, yet iron can still tell them apart.
Xpeng says this works because iron captures facial features and voice print information
when meeting someone, then connects those signals with details from their conversation.
Together, this becomes what Xpeng describes as a dedicated digital memory profile.
When iron encounters that person again, it can retrieve the profile and continue updating it.
That is why iron recognizes Xiaohei and remembers that he has three children
and was interested in the G9L for camping.
The robot is connecting an identity with a history of previous conversations,
rather than simply remembering a face.
Xpeng calls this multimodal recognition because iron can combine face,
voice, and conversation history instead of depending on one clue,
changing a shirt, or moving somewhere else in the room,
therefore does not automatically erase the information the robot has connected to that person.
The conversation also moves between English,
Chinese, and German without anyone telling iron to switch modes.
Xpeng says this comes from a multilingual foundation model
combined with post-training and balanced multilingual training data.
The goal is to make language switching part of the conversation itself.
When someone changes languages,
iron can follow that change and answer in the language it hears.
What if you walked into a showroom six months later
and iron still remembered your name and exactly which car you wanted?
Would that impress you or feel strange?
Comment below!
All of these technologies make more sense when you look at the job Xpeng is preparing iron to do.
The company says one of its early roles could be working as a guide inside Xpeng stores,
where remembering customers and understanding what they need
becomes much more useful than simply answering generic questions.
When Xiaohei asks whether the G9L can carry his family of five along with their filming equipment,
iron already has context from their earlier conversation.
It combines that memory with vehicle information
and explains that the G9L offers more than 1,100 liters of luggage space.
Xpeng's published figure is 1,152 liters.
The intended workflow is simple.
Recognize the customer, retrieve what they need,
and connect that information with relevant product knowledge.
Powering these capabilities are three of Xpeng's own Turing AI chips
with up to 2,250 tops of effective compute.
Xpeng says iron can use that onboard processing to access a local knowledge base with low latency,
while online retrieval can provide information when needed.
This gives the robot substantial computing power directly inside its body,
rather than making every interaction dependent on a remote computer.
A showroom gives Xpeng a practical place to bring all of this together.
Iron can recognize visitors and remember their preferences
while answering questions about a controlled range of products.
Xpeng is targeting production by the end of 2026 with deliveries planned for 2027.
While Xpeng is teaching a humanoid how to understand the people around it,
Pittsburgh-based Skilled AI is exploring a very different kind of learning.
Its latest demonstration puts a Unitry G1 humanoid on a football pitch against real people.
Skilled provides the AI controlling the robot,
and the footage shows it operating autonomously at normal speed while wearing an Argentina jersey.
The company calls it the Messinator.
The interesting part is not simply that the robot can kick a ball.
It dribbles around defenders, protects possession with its body, and makes tackles.
Those abilities came from a training process that combined basic human guidance
with an enormous amount of simulated experience.
Training began with human-referenced drills for basic skills, such as kicking and dribbling.
Skilled then moved the AI into NVIDIA's Isaac Sim and used self-play.
The robot repeatedly competed against recent versions of itself,
which meant every improvement created a stronger opponent
and forced the newer version to find better strategies.
Simulation allowed this experience to accumulate
far faster than practicing with one physical robot.
Skilled says the process added up to around 140 years of simulated football experience.
Early in training, the humanoid could barely walk.
With more experience, it learned to recover from falls,
while behaviors such as shielding the ball and tackling
began appearing because those actions helped it score.
Humans provided the basic football foundation,
but they did not need to explicitly tell the robot
to turn its body between the defender and the ball.
Strategies like that emerged through self-play,
because the robot discovered which movements helped it reach its objective.
Skilled eventually transferred this learned policy from simulation
into the physical Unitry G1.
This step matters because virtual physics can never perfectly reproduce a real robot,
floor or football.
Yet the behaviors learned in simulation carried into the physical matches shown by Skilled.
During operation, the controller sends new targets to the robot's joints 50 times every second,
so it can continuously adjust its movement and balance.
Would you play one-on-one against this robot
knowing it has 140 years of simulated football experience?
Comment below!
Football is really a test bed for something much larger.
Skilled wants to use physical self-play to give robots enormous amounts of experience
before bringing those learned behaviors into factories, construction sites, and homes.
Thousands of simulated robots could practice at the same time
in situations where one physical robot would need years to collect the same amount of experience.
And when you look at both robots together,
humanoid development is clearly moving beyond simply making a robot walk
or perform a prepared routine.
Xpeng is working on perception, memory, and natural interaction,
while Skilled is exploring how robots can develop physical strategies through experience.
And if you want to see where these robots go next before they become part of everyday life,
subscribe to AI Nexus.
We track the breakthroughs that are actually pushing humanoid robots forward.
These are very different approaches,
but both point toward humanoids that can understand a changing situation
and adapt their response instead of following one fixed routine.
And as companies continue solving different pieces of that problem,
we are starting to see what happens when humanoid robots are given not just better bodies,
but better ways to understand and learn from the world around them.
Agility Robotics may have just changed the humanoid game again,
and this time it is not about another viral robot demo.
Its new Digit 5 is built around one problem factories cannot ignore,
how a large humanoid can safely work around people.
The robot can detect when someone gets too close, then avoid them,
stop moving, or even lower itself into a stable seated position.
It can repeatedly carry up to 50 pounds and squat much deeper than Digit 4.
Its reach has jumped to more than 7 feet.
Its battery only runs for around 90 minutes,
but Agility says it can recharge in just 9 minutes.
And instead of giving Digit 1 permanent hand,
the robot can use different tools depending on the job.
Even its famous backward-looking legs have been redesigned.
So this is not just Digit 4 with a few upgrades.
Agility has rebuilt the robot around safety, heavier work, and much longer daily operation.
Digit 5 is Agility Robotics' 5th generation humanoid,
and the company is positioning it very differently from many humanoids we usually cover.
This is not being presented mainly as a research platform.
Agility wants it working inside warehouses,
distribution centers, factories, and logistics sites.
The design came from roughly 3 years of feedback collected
while Digit 4 was operating inside real customer facilities.
Digit 4 already proved that a humanoid could do repetitive commercial work,
but there was still a major restriction.
The robot usually worked inside controlled areas that kept ordinary workers separated from it.
Digit 5 is Agility's attempt to move beyond that model.
Agility calls the idea Cooperative Safety,
where dedicated safety hardware helps the robot operate closer to people.
Digit 5 stands 5 feet 11 inches tall and weighs about 284 pounds.
Its height is human-like, but it is much heavier than a person.
Agility still wants it to fit through normal doors,
move through human-sized aisles, and use workstations built for people.
The first big hardware upgrade is Payload.
Digit 4 was rated for up to 35 pounds.
Digit 5 raises that to 50 pounds for repeated material handling, roughly a 40% increase.
This is not a one-time maximum lift.
The robot is being designed to move those loads repeatedly during an industrial shift.
The legs may be an even bigger change.
Older Digit robots were known for their backward-looking leg design,
which came from Agility's earlier Cassie robot.
Those legs were excellent for dynamic walking,
but they were not ideal when the robot needed to get very low to the floor.
Digit 5 now uses a more human-like knee arrangement.
And the goal is not simply to make the robot look more human.
The new geometry lets Digit squat deeper and place its body in a better position
when lifting from low shelves, pallets, or containers near the ground.
Agility also built its own cycloidal actuators into the new design.
These joints are meant to handle repeated heavy-duty motion.
The focus is what happens after the robot has lifted
the same kind of load hundreds or thousands of times.
But the biggest headline is safety.
Digit 5 includes what Agility calls safe human detection.
Multiple sensors and a separate safety layer
keep checking the area around the robot for people.
If someone enters a risky distance,
Digit does not have to rely only on its normal AI to decide what happens next.
An independent safety controller can trigger a response immediately.
Depending on the situation, Digit may avoid the person,
stop moving, or lower itself into a seated position.
That seated behavior is especially interesting because it places the robot
into a more stable state instead of forcing a 284-pound humanoid
to keep dynamically balancing while someone is close.
Agility has also added visual and audio signals
so people nearby can better understand what the robot is about to do.
The company is also working with NVIDIA through its Halos for Robotics platform
using NVIDIA IGX-Thor and Halos Core as part of the wider safety architecture.
And this safety work is not starting from zero.
Agility says it is building on more than 65,000 hours
of operating experience gathered from previous Digit deployments.
If you want to keep up with humanoid robots
that are moving from demos into real industrial work,
subscribe to the AI Nexus because the next change
is one of the most unusual decisions Agility made with Digit 5.
The battery actually looks worse at first.
Digit 4 was advertised with around 4 hours of runtime,
while Digit 5 is designed for about 90 minutes.
But Agility completely changed the charging strategy.
Digit 5 can recharge in around 9 minutes,
giving it roughly a 10-to-1 operating-to-charging ratio.
That means the robot does not need one huge battery session.
It can work for about an hour and a half, recharge quickly, and then go back to work.
Agility says this approach could give Digit 5
more than 20 productive hours during a 24-hour day.
Charging is also autonomous.
When the battery gets low, Digit is designed to return to its charger
without waiting for an employee to plug it in.
That becomes much more important if a warehouse eventually has dozens
or even hundreds of these robots operating across several shifts.
Reach has improved too.
Digit 4 could reach roughly 5.5 feet vertically.
Digit 5 can reach around 7.2 feet.
That gives the robot access to much higher shelves and equipment
without forcing companies to redesign everything around it.
Agility has also moved away from the idea that one human-like hand should do every job.
Digit 5 uses an ISO standard mounting point for interchangeable end effectors.
In simple terms, different grippers or tools can be attached depending on the task.
Agility has shown research involving five-finger robotic hands.
For industrial work, Agility is currently choosing simpler tools that can be strong
and reliable instead of trying to copy the appearance of a human hand.
And that connects directly to the biggest change in what Digit is supposed to do.
Digit 4 became known mainly for tote movement.
It could pick up containers, carry them between stations,
work around conveyors, and place those containers somewhere else.
That proved the basic idea,
but it still meant the robot was doing a fairly narrow part of the overall workflow.
Digit 5 is supposed to go much further.
Agility lists depalletizing, machine tending,
kitting, sequencing, quality inspection, material movement, and palletizing as target jobs.
Instead of buying a humanoid only to move the same tote between two points,
the idea is that the same Digit platform could
eventually perform several different jobs inside one facility.
Machine tending could be especially important.
Digit now has more reach, more payload, a deeper squat, and interchangeable tools.
That combination could let the robot take material from a container,
move toward a machine, load or remove a component,
and then carry that material to the next part of the process.
Behind all of this is Agility's physical AI stack.
The company divides the intelligence into two broad areas.
Semantic AI handles things like understanding a task,
planning, and reasoning about what needs to happen.
Physical AI deals with actually doing it.
That includes walking, balancing, controlling force,
manipulating objects, and coordinating the entire body.
Agility also uses large language models and vision language action models in that stack.
So Digit 5 is not the first intelligent Digit.
The bigger change is how that AI is being combined with new hardware,
new safety systems, and a body designed for wider industrial use.
And Agility has something many humanoid companies still do not have.
A large amount of real deployment experience.
Previous Digit robots have accumulated more than 65,000 operating hours.
At GXO's Flowery Branch facility near Atlanta,
Digit also passed 100,000 totes moved.
Agility says the robot maintained around 98% accuracy
while it was actively performing that work.
Those deployments matter because Digit 5 was not designed only from simulation or lab testing.
Agility says its training process combines real customer data,
motion capture, teleoperation, simulation, and repeated practice.
Digit 5 will also connect to Agility Arc, the company's fleet management platform.
Arc can assign work, monitor performance,
and connect Digit with warehouse software, conveyors, and autonomous mobile robots.
That lets Digit become one part of a larger automated workflow instead of operating alone.
But there is one important reality check here.
Digit 5 has been unveiled, but it is not yet shipping at large scale.
Agility expects early access during the first half of 2027,
followed by broader availability by the end of that year.
The company also plans to expand beyond the United States and Canada
into Europe and the United Kingdom.
Agility says it already has more than $300 million in multi-year Digit 5 orders,
although those orders depend on product and contract milestones.
Investor Material connects that commitment with roughly 1,000 robots
under a three-year Robots-as-a-Service contract.
So Agility has not already earned that entire $300 million,
but customers are clearly planning around the new platform.
Agility is also preparing for scale at RoboFab in Salem, Oregon.
The humanoid factory is designed for peak output of up to 10,000 Digit robots per year,
although actual production will depend on demand.
And that is what makes Digit 5 different from a normal generational update.
Digit 4 proved that a humanoid could keep doing a real industrial job for long periods.
Digit 5 is Agility's attempt to take everything learned from that experience
and remove some of the biggest limits around it.
The robot can carry more, squat lower, reach higher, recharge far faster,
switch tools, and cover more jobs.
But the biggest change is still the safety architecture.
If Agility can make this kind of cooperative human robot work reliable in real facilities,
Digit could spend far less of its working life behind barriers
and far more of it moving through the same workplaces built for people.
