Bespoke Robot Society:Battlebots Are Not Robots
BattleBots Are Not Robots
This is the first BRS rant, addressing a fundamental misconception about what constitutes a robot.
The Problem
Everyone hears the word "robot" and they immediately jump to "oh, let's make a BattleBot." This popular show has been in the zeitgeist for possibly decades now—I think it's still going on. I face this problem constantly: you say "I'm building a robot," and everyone's like "oh, I've heard of BattleBots, I know all about those."
This is such a poisonous understanding of what a robot is.
What Is a Robot?
So first of all, what's the definition of a robot?
A robot is something that has a control loop behind it. It's software reading data from sensors, making decisions, and taking actions. It's about:
- Navigating
- Perceiving
- Manipulating the world
Why BattleBots Aren't Robots
A BattleBot is not a robot. A BattleBot is just remote control. The BattleBot itself is only a radio receiver and a bunch of actuators. It's a puppet. There's no robotics to it because it's not using any sensors, it's not deciding what to do—that's just a person driving it. Not even a robot. The "bot" part of it is just a lie. It's just battle.
See also: Beyond Toys for more on the distinction between robots and remote-control toys.
What Robotics Is Really About
Philosophically, what is a BattleBot in comparison to the field of robotics? It doesn't automate anything. It doesn't solve a problem. It doesn't make your life easier or better. It doesn't enable some process that wasn't possible on human labor alone.
It's only destruction entertainment. They're completely aware of this, by the way they market it and design the show. This is no different from a destruction derby or those mud pits where they crash cars into each other. There's nothing robotic about the entertainment value of that show.
Actual robotics is about:
- Problem solving
- Innovation
- Reducing cost
- Accuracy, repetition, and perfection
It's not about just smashing things by remote control. I think BattleBots has completely associated the wrong ideas with "robot" in people's minds.
The BRS Alternative
BRS's mission, specifically, is supposed to get away from the BattleBot thing. BattleBots are remote control toys. I want all of Bespoke Robot Society's robots to serve an actual purpose.
Now, the purpose might just be that you don't know that much about robots and you want to fix that in a hands-on way. That's a problem that a BRS robot design could solve. I'm completely happy with that being the problem. But smashing up another robot? Not really an interesting problem I want to solve with BRS robots.
Autonomy on a Budget
The very first Bespoke Robot Society robot—SimpleBot—already has more sensors than the most sophisticated BattleBots, and it's only got line sensors.
A BRS robot is about demonstrating that loop even under $20:
- Have sensors
- Make a decision based on those sensors
- Actuate in the world—use some kind of manipulator
SimpleBot already has all those elements:
- It detects lines on the ground
- It has programmed logic to respond to those lines
- It changes its behavior in the world
This is actual robotics, not just remote-control toys. See Toddler Testing for how we validate that these robots work in the real world.
Battle Without Destruction
Now, will we play war games? Sure. One of my prototypes here, the Tiny Tank 2, has a pan-tilt camera, a front-facing camera, and a LiDAR sensor. The intended use case for that robot is backyard-scale autonomous navigation and remote operation—either two-person (one driver, one gunner) or one of those roles, or both of those roles being served by AI—to interact with each other on the battlefield of my grassy yard.
We'll use IR blasters pointed at each other, detecting being hit, detecting that there's a signal coming at you. That's a way to have some of the tactical excitement of a BattleBot without being focused on destruction. And it's also using it as an entry point to artificial intelligence and autonomy of the robot—whether that robot is literally learning by playing against us, or we custom-write software as an additional challenge for us, the designers, to try to write software ourselves that solves the problem rather than just remote-controlling it and solving the problem one time.
You want to write software that solves it every time. Those are the sorts of things I think actually embody the spirit of robotics—the direction I want to take Bespoke Robot Society in.