1. Global Robotics & Science Foundation
  2. General
  3. General
  4. Other Resources

Design Convergence: Why Robots Look Alike

Go to the first event of the season and look around. The robot designs are mostly unique, and it's hard to find two that look alike. Walk the pits at any event in the second half of the season and the picture has changed. Most robots have settled on a few common designs. Drive bases use the same few layouts. Intakes work the same way. Lifts follow a small number of designs that have been proven to work. By the VEX Robotics World Championship, the robots on the field often have more in common than not.

Every season, somebody looks at all that similarity and cries foul. A judge gets suspicious, a parent posts two photos side by side, or a team gets accused of copying because its robot looks like one from another team.

This article explains why similarity happens, because it causes more confusion and unfair suspicion than almost anything else in competitive robotics. Designs grow more alike whenever many people work on the same problem under the same rules. We call the similarity Design Convergence. When you see it, it means a lot of teams are doing real engineering on the same problem, and it is not evidence of cheating. This article is the reasoning behind the "When designs look alike" section of the Student-Centered Policy, and is written for judges and coaches.

Why Designs Converge

Four things push designs together, and all four start the day VEX releases the game.

  • The game limits the options. It says what the robot has to do to earn points, how big it can be, which parts are allowed, and how many weeks you have to build. Before anyone has cut a single piece of metal, those limits have already ruled out most of the ways a robot could be built. What's left is a short list of designs that can actually work, and every team in the world is choosing from that same short list.
  • Teams can see each other. Reveal videos and match footage show which mechanisms work and which ones break. Photos get shared, students talk about what they saw last weekend, and by December most teams know which drive bases and intakes are winning. A team that uses that knowledge isn't cheating. Working engineers read what other engineers have published before they design anything, and a team that studies other robots is doing the same thing.
  • Teams help each other on purpose. They scrimmage together, mentor rookie teams, post their builds, and answer questions in the pits. Part of that comes from how the program is set up: elimination rounds are played in alliances, so it pays to have strong teams around you. Most of it is just what people do when they gather around the same problem in the same gym, with every robot out in the open. No rulebook could stop that teaching, and ours doesn't try, because the program celebrates teams that make other teams better.
  • Losing designs disappear. Other teams study the mechanisms that keep winning and improve on them. The ones that keep failing get taken apart and rebuilt as something else. Repeat that at thousands of events and hundreds of thousands of matches. The designs still on the field at VEX Worlds are the ones that kept working against real opponents.

Put the four together and you get convergence, which is simply what engineering looks like when many people do it at once. Evolution runs the same loop, and nature needs millions of years for it. A robotics season does it in one.

Three Kinds of Honest Resemblance

When one team's robot looks like another team's, three honest explanations cover almost every case.

  • Convergent evolution —Two teams reach the same solution separately because the game pushes everyone toward it. The physics made the answer available to anyone who thought carefully about the problem, so nobody copied anybody and no team owns the idea.
  • Inspired adaptation — A team sees an idea on another robot, watches a reveal video, or has a conversation with a coach. They learn what makes it work, prototype their own version, test it, and build it into a robot they understand. The inspiration came from outside, and the engineering is theirs. This is how the engineering profession works, whether it's a working system from a textbook, working code from a library, or a working product already on the market. It's also how people learn. Research on learning going back to the 1980s shows that beginners learn faster when they study worked examples before trying problems on their own. They also apply what they learned more easily. A student who studies another team's mechanism and then engineers their own version is learning the way the research says people learn, not cutting corners.
  • Shared work — Teams trade ideas, teach each other, and work through problems together, and two teams that collaborate can show up with robots that look alike. That isn't copying as long as each team did its own work, tested the result, credited the collaboration, and can explain its own robot. 

Asking the Right Question

The right question to ask is, Did this team do the engineering work for this design? as opposed to Is this robot similar to that one? The Student-Centered Policy turns the right question into three Checks, summarized here: 

  1. Student-Centered Check: Are we doing the work? 
  2. Help Check: When we get help, does it teach us?
  3. Ownership Check: When we use an outside idea, do we make it our own? 

Yes to all three, means the team is student-centered.

This article is about how teams use an outside idea and show that they make it their own. The Ownership Check breaks that into four parts: Adapt, Test, Understand, Credit. If the answer to all four is yes, the idea is now the team's own. If any of them is No, the team has more work to do to make the idea their own. 

What Design Convergence Is Not

Resemblance stops being honest when the team did not do the designing. That happens two ways: 

  • Direct copying — A team rebuilds someone else's design down to the details without understanding why the choices were made, without testing for improvements, without doing the engineering that turns an idea into a working system of its own. The hands were the team's. The work was not.
  • Purchased work — Outside ideas can always be made your own; finished work cannot. Bought designs, bought code, and robots handed over as favors arrive with the engineering already done, so there is nothing left to make your own. Paying to learn is normal: college, a robotics camp, a class on any subject. Paying to leap past the learning fails, because the learning is the point.

Both end the same place: the team is claiming work it did not do, and that claim is the integrity concern. Neither is design convergence, because nothing converged. One design was taken, the other was bought. Neither is student-centered.

What This Means for Judges

Judges should expect to see similar designs at a competition, especially late in the season. They should ask questions about that work like:

  • Did you make the decision on what to build?
  • When you received help, did it teach you?
  • Where did outside inspiration come from? 
  • What did you keep, change, or throw out, and why? 
  • Can you walk through the mechanism and explain how it works? 
  • Does the notebook show the journey?

A team that can answer those questions has done student-centered engineering, however much its robot looks like everyone else's. A team that can't has not done the work, however original its robot looks. What should concern a judge is missing work or hidden sources, never resemblance by itself.

What This Means for Coaches

When your team's robot starts looking like everyone else's, that is the season working, not a problem to fix. Say it out loud: students hear the copying accusations too, and they need to know that resemblance is normal. Send them to the reveal videos, the pits, and the match footage, because studying other robots is part of the work. 

  • When an outside idea comes home, coach the four steps of the Ownership Check that make it theirs: adapt it, test it, understand it, and credit it. 
  • Insist on the credit the day the idea arrives, while everyone still remembers where it came from. 
  • If your team is ever accused because its robot resembles another, the answer is the team's own record: the notebook, the credits, and students who can explain every mechanism.

Convergence Outside of Robotics

Design convergence happens everywhere; not just in robotics. If robots were the only things that converged, you could argue robotics teams have a copying problem. But the same thing happens when many separate groups work on the same problem under the same limits, and the examples come from many different places.

Good designs dominate. There is a reason almost all cars work the same way. The steering wheel, the pedal layout, the basic shape: these are the ideas that won after the bad ideas died away. As long as no patent is violated, adopting them is perfectly legal. The same is true on the kitchen counter. Two toasters from two brands have different controls and do the identical job, and the toaster settled into two main designs, top-load and front-door load. Innovation researchers William Abernathy and James Utterback traced this pattern in the 1970s. The researchers now call it a dominant design: the layout that wins over an industry and becomes the starting point every competitor builds on. The DC-3 did that for airliners. A robotics season runs that whole industry cycle in a single year.

image2.jpg
The Brawn BGP 001, Abu Dhabi 2009: the double-diffuser car three teams reached independently. Photo: Yuriy Lapitskiy via Wikimedia Commons, CC BY-SA 3.0.

Racing is like robotics. 

Formula 1 is the closest professional comparison to what our teams do: a fixed rulebook, a fixed season, and as much cleverness as a team can bring. In 2009, three teams (Brawn, Williams, and Toyota) showed up with the same "double diffuser." Each had found the same gap in that year's new aerodynamic rules on its own. Once it was ruled legal, every other team on the grid built one as fast as it could. Both halves of that story are normal: three teams finding the same idea on their own, and everyone else adapting it the moment they saw it. Nobody in racing calls either half cheating.

CleanShot 2026-09-16 at 15.33.38.png
February 14, 1876: Bell's patent (left) and Gray's caveat (right), filed hours apart. Public domain: USPTO, 1876 (left, right).

One invention, multiple inventors. 

On February 14, 1876, Alexander Graham Bell's lawyer filed papers at the US Patent Office for sending speech by electricity. A few hours later that same day, Elisha Gray filed his own papers for a device meant to do the same thing, based on his own design. That case is famous, but it isn't the exception, it's the norm. Stanford law professor Mark Lemley looked at the history of major inventions, from the light bulb to the airplane. Almost all of them, he found, were invented at nearly the same time by two or more people working separately. A big part of patent law exists to sort out exactly that situation.

image5.png
A Jurassic ichthyosaur and a modern dolphin, roughly 200 million years apart. Reptile and mammal, never alive at the same time, and the ocean shaped both the same way. Diagram: Sceptic view via Wikimedia Commons, CC BY-SA 4.0.

Nature converges. 

A shark is a fish, an ichthyosaur was a reptile, and a dolphin is a mammal. Nobody copied anybody. The ocean gave all three the same problem, moving fast through water, and all three ended up with the same sleek shape. Octopus eyes and human eyes work the same way too, and they developed separately. Biologists call this convergent evolution, and it happens with no communication at all: the same problem produces the same answer. A robot game field poses a problem the same way an ocean does.

image3.jpg
The flop, decades on: Chaunté Lowe, Karlsruhe 2010, now the standard technique at every level. Photo: Grzegorz Jereczek via Wikimedia Commons, CC BY-SA 2.0.

Humans converge too. 

Dick Fosbury won high-jump gold in 1968 by going over the bar backward. Four years later in Munich, 28 of the 40 jumpers in the men's event were using his technique. They watched a better way to jump, understood it, and trained it into their own form, and sport celebrated them for it rather than calling it theft.
 

Do All Designs Converge? 

No. Convergence is a tendency, not a law. Early-season VEX events are full of unique designs, and new approaches keep showing up all the way to VEX Worlds. Some teams resist the norm on purpose, either as part of their identity or because they believe their design has a real chance. Those unique designs often end up winning Design Awards. 

Even after most of the field has converged, a different method can still win. The 1972 Munich men's gold went to a jumper using the older technique, not Fosbury's. Fosbury's way was better for most jumpers, but convergence describes the majority, not who wins any single match. Skill still decides that. 

Nobody is saying every team must converge. The point is that when it happens, it isn't evidence that anyone did something wrong. Design Convergence is what happens when many teams work toward the same goal. We should recognize it and welcome it. And we should expect every team to be honest about where its ideas came from, because that honesty is what separates engineering from copying.

Find ideas. Make them yours. Have fun.

Credit Summary

This article used outside ideas too. Here they are, gathered in one place.

Last Updated: