SupplyChainToday.com

Robot Fight Clubs May Be Building the Humanoids That Transform Supply Chain.

At first glance, humanoid robots punching and kicking each other inside a ring can look like futuristic entertainment. Watch the NBC News segment “Inside the world of robot fight clubs,” however, and a much bigger story begins to emerge. The segment features humanoid robotics industry analyst Eren Chen discussing the growing robot-fighting trend in China. The obvious attraction is the spectacle. The less obvious story is the engineering happening behind it.

A robot that can fight has to solve some of the hardest problems in robotics at the same time. It must remain balanced while moving quickly, understand where another moving object is located, coordinate its arms, legs and torso, react rapidly, absorb physical impact, recover from mistakes and continue operating when conditions become unpredictable. Those are fighting skills inside the ring. In a factory or distribution center, many of those same abilities become valuable business capabilities. Research into humanoid fighting is already focusing on stability, whole-body control and smooth transitions between complex movements.

That is why robot fighting may matter much more to supply chain professionals than it first appears.

Think of it somewhat like motorsports and the automobile industry. Racing is not normal driving, but extreme competition creates an environment where equipment is pushed toward its limits. Weaknesses become obvious very quickly. Engineers learn what breaks, what survives and what needs to improve. Robot fighting could play a similar role for humanoid robotics. The ring becomes a brutal testing laboratory where balance, sensors, actuators, software, batteries, joints and control systems are pushed far beyond the comfortable conditions of a carefully staged demonstration.

The Fight Is Really About Balance, Perception and Recovery

One of the most striking examples came from the Ultimate Robot Knock-out Legend competition in Shenzhen, China. During a match, an EngineAI T800 humanoid called White Eagle delivered a high kick that knocked the head from another robot, Matador. Remarkably, Matador continued operating using systems in its torso. The competition evaluates more than striking power; the robots are tested on stability, defensive movement, agility, durability and their ability to recover from falls.

Now forget about fighting for a moment and translate those capabilities into a warehouse.

A warehouse robot may eventually need to step around a misplaced carton, recover after bumping an object, pull a cart whose load suddenly shifts, lift something awkward without losing its balance, or continue operating when a sensor is partially obstructed. A manufacturing humanoid may have to reposition its feet while handling a component, work around another machine or immediately stop when a person unexpectedly enters its path.

The more dynamic an environment becomes, the less useful a robot is if it only knows how to perform one perfectly scripted motion.

Traditional industrial robots are extraordinarily productive when the environment around them is controlled. Bolt a robot arm to the floor, put the same part in the same position and repeat the same movement thousands of times, and automation can work beautifully. Supply chains are different. Boxes arrive damaged. Inventory gets placed in the wrong location. Product sizes vary. People walk through the work area. Priorities change. Demand changes.

Humanoid robots are being developed specifically around the idea of functioning in environments originally designed for humans. Companies including Agility Robotics and Apptronik describe their humanoids as machines intended to operate in existing warehouses, factories and other human workspaces.

From the Fighting Ring to the Warehouse

The business case for humanoid robots is not simply that they look like people. It is that much of the industrial world was already designed around the human body.

We have stairs, doors, shelving, carts, workstations, tools, conveyors and aisles designed around what people can reach, carry, push, pull and navigate.

If a robot can move through those environments, manipulate objects and learn additional tasks without requiring a company to completely redesign its facility, automation becomes far more flexible.

We are already seeing early examples of that transition. Agility Robotics says its Digit humanoid entered commercial operations at a GXO facility in 2024. There, Digit works alongside existing automation, including autonomous mobile robots, moving totes from those robots onto conveyors. By November 2025, Agility reported that Digit had moved more than 100,000 totes at the facility.

Figure has also put humanoids into automotive manufacturing. The company reported that Figure 02 accumulated more than 1,250 operating hours and contributed to production involving more than 30,000 BMW X3 vehicles before that generation was retired. In June 2026, Figure demonstrated its newer Figure 03 performing a logistics workflow at BMW Group Plant Spartanburg that combined manipulation, walking, repositioning its body and pulling a wheeled cart.

Apptronik is pursuing similar opportunities with its Apollo humanoid. The company describes applications including moving components, sorting, kitting, inspection and machine tending, and it has established relationships with companies including Mercedes-Benz and GXO.

None of this means humanoids are ready to walk into every warehouse tomorrow. It does mean the conversation has moved beyond science-fiction speculation.

Why Fighting Could Accelerate Industrial Robots

Robot fighting could accelerate progress for a simple reason: failure is immediate and obvious.

In a carefully controlled demonstration, engineers know much of what is supposed to happen. A robot can perform a choreographed movement and everyone applauds. Put two humanoids into a fighting ring and much of that predictability disappears.

The robot gets pushed into positions engineers may not have anticipated. Its center of gravity changes rapidly. Sensors must interpret fast movement. Software has to transition between walking, turning, defending, striking and recovering without the robot falling over.

Researchers are already working directly on these challenges. A 2026 research paper on humanoid fighting examined how robots can transition smoothly among multiple movement skills while maintaining stability. Another 2026 project called HumanX demonstrated a system that learned several real-world interaction skills—including cargo pickup and reactive fighting—and transferred those skills to a physical Unitree G1 humanoid.

Think about that connection for a moment.

Cargo pickup and robot fighting sound like completely different activities. From the robot’s perspective, however, both require perception, movement, balance, timing, coordination and interaction with a changing physical environment.

That is the bigger story. The better robots become at general physical intelligence, the more tasks a single machine could eventually perform. And that could change the economics of supply chain automation.

Today, companies frequently automate processes by purchasing specialized equipment for specific jobs. One system moves pallets. Another sorts cartons. Another transfers totes. Specialized automation will remain extremely valuable, especially for repetitive, high-volume processes.

But a capable humanoid introduces a different possibility: a general-purpose machine that could eventually be reassigned as business requirements change. Imagine a humanoid unloading material during one part of a shift, replenishing an assembly line later and assisting with finished-goods movement afterward. We are not fully there yet. But that is the direction that makes humanoid robotics so strategically interesting.

Physical AI May Be the Next Layer of Supply Chain Automation

For the last several years, much of the attention around artificial intelligence has focused on information. AI can analyze data, identify patterns, assist with forecasting, create reports and support better decisions.

Humanoid robots bring AI into the physical world. The system does not simply recommend that a box move from Point A to Point B. The robot sees the box, walks toward it, determines how to grasp it, positions its body, picks it up, maintains its balance, carries the box and puts it where it belongs. That sounds easy because a person can do it almost without thinking. For a robot, it is enormously complicated.

This is why advances in whole-body control are so important. Figure’s Helix 02, for example, is designed to coordinate the robot’s hands, arms, torso and feet together during longer autonomous tasks. Figure has demonstrated the system performing extended tasks requiring both locomotion and manipulation rather than treating walking and object handling as completely separate activities.

The important shift may eventually be from automation that follows a programmed route to automation that understands a task. That is a much bigger idea.

What Supply Chain Leaders Should Do Now

The wrong response to an impressive humanoid video is immediately asking, “How many robots should we buy?”

Start with the work.

Walk through your warehouse, distribution center or factory and identify jobs that are repetitive, physically demanding, ergonomically difficult, hard to staff or located in places where traditional automation doesn’t make economic sense.

Pay particular attention to the gaps between automation systems.

Those gaps are often where people are still moving totes, carts, containers and components because conventional automation is too rigid or too expensive to extend. Digit’s commercial work at GXO offers an interesting example: the humanoid helps connect an existing automated workflow by transferring totes from mobile robots to a conveyor.

Then ask a better question: What capabilities would a robot need to perform this task safely, reliably and economically? Would it need dexterous hands? How much weight would it carry? Would it need to work around people? Pull carts? Handle mixed products? Recover when something isn’t exactly where it is supposed to be? Connect with a warehouse execution or management system?

That changes humanoid robotics from a technology conversation into an operations conversation. Leaders should also resist evaluating robots based on viral demonstrations alone. A backflip is impressive. A spinning kick gets attention. But businesses don’t make money from backflips. They make money from uptime, throughput, quality, safety, flexibility and cost.

The humanoid robot that ultimately wins in supply chain may not be the one with the most spectacular demonstration. It may be the one that quietly shows up for thousands of shifts, handles exceptions safely, integrates with existing automation and produces a measurable return on investment.

The Real Lesson Behind Robot Fighting

Robot fight clubs are easy to dismiss as a novelty.

That may be a mistake.

The fighting ring is exposing humanoid robots to the type of unpredictability engineers must eventually solve if these machines are going to become truly useful outside highly controlled environments. Every fall tests recovery. Every impact tests durability. Every evasive movement tests perception. Every transition tests software coordination. Competition can create information about what works—and what breaks. EngineAI’s leadership has explicitly described its fighting competition as a way to help drive robotics research and industrial development.

The fighting itself isn’t the destination.

Better physical intelligence is the destination.

And supply chain could become one of its biggest beneficiaries.

Factories and warehouses contain enormous numbers of jobs that still depend upon human mobility, dexterity and adaptability. As humanoid robots become more capable, the boundary between what can and cannot reasonably be automated will continue to move. Commercial deployments by GXO, BMW and others suggest that this transition has already begun, even though the technology still has a long way to go.

The companies that benefit most may not be the ones that buy humanoid robots first.

They may be the companies that understand their processes well enough to recognize exactly where these machines can create value, redesign work intelligently and combine human strengths with machine capabilities.

So watch the robots fight and enjoy the spectacle.

But look past the punches.

You may be watching the training ground for the next generation of supply chain automation.

Want to stay ahead in the supply chain game? Subscribe to our newsletter for the latest trends, insights, and strategies to optimize your supply chain operations.

History of Robot Fighting

  • 1994 – Robot Wars begins modern robot combat. Designer Marc Thorpe organized the first Robot Wars competition in San Francisco, where remote-controlled machines battled in an arena. It helped establish robot fighting as both an engineering competition and spectator sport.
  • Late 1990s – Robot fighting becomes television entertainment. The British Robot Wars television series launched in 1998, bringing destructive robotic battles to a mass audience and inspiring builders around the world.
  • 2000s – BattleBots expands robot combat in the United States. BattleBots helped turn combat robotics into a recognizable competitive sport, with purpose-built machines using spinning weapons, hammers, flippers and other designs to disable opponents.
  • 2002 – Humanoid robot fighting takes a major step forward. Japan held the first ROBO-ONE competition, featuring small bipedal humanoid robots fighting one-on-one. Unlike traditional wheeled combat robots, these machines had to walk, balance, strike and recover from falls.
  • 2017 – Humanoid fighting begins moving toward autonomy. ROBO-ONE introduced an autonomous biped fighting tournament, reducing the role of human operators and requiring robots to sense, decide and fight more independently.
  • 2025 – Full-size humanoid fighting enters a new era. China hosted highly publicized humanoid kickboxing competitions, demonstrating punches, kicks, balance and increasingly sophisticated whole-body control. These events began looking less like traditional remote-controlled robot combat and more like testing grounds for advanced humanoid robotics.
  • 2026 – Robot fighting becomes a serious physical-AI test bed. Competitions such as Shenzhen’s Ultimate Robot Knock-out Legend pushed larger humanoid robots through high-impact fighting, testing stability, perception, durability and recovery. At the same time, robotics researchers are developing AI systems that allow humanoids to transition smoothly among walking, running and fighting skills—capabilities that could eventually transfer to factories, warehouses and logistics operations.

Robot Resources

1 2 3

Leave a Comment

Scroll to Top