Robotics Is in Its 1973 Era | Evan Beard artwork

Robotics Is in Its 1973 Era | Evan Beard

MTS

October 28, 2026

Standard Bots co-founder and CEO Evan Beard breaks down why robotics is at its "1973 computer" inflection point.

0:00

Evan Beard

The broader context in robotics and manufacturing is 6% of manufacturers have tried a robot at all. So all these things are really two axes, which we think can move that 6% number. One is making the robot so anyone can take them out of the box, like an iPhone type experience for what traditionally is more of a mainframe computer type product. And the other is making them more capable, so you can just automate a lot of jobs that previously had too much variability and you couldn't automate.

0:24

SPEAKER_2

Hello everyone, welcome back to MTS. We are joined with Evan Beard. He is the co-founder, CEO and chief engineer of Standard Bots. Standard Bots is America's largest manufacturer of AI native industrial robots. And they're joining us today on MTS to talk about all of their latest updates and announcements. So Evan, welcome to MTS.

0:44

Evan Beard

Hey, thanks so much for having me. Really appreciate it.

0:46

SPEAKER_2

Thank you for joining. I'm very excited to have you. I know you guys had some big announcements this October. You nonced the Standard Bots AI, the AI locate skill. What's the latest right now happening with Standard Bots?

0:58

Evan Beard

Yeah. Well, the broader context in robotics and manufacturing is 6% of manufacturers have tried a robot at all. And so we're releasing a series of new features and products that really make it so that someone can take a robot out of the box without having any experience in robotics and install that robot and get it running. And maybe the only graduated community college or high school, and maybe you're working on a manufacturing floor and that's your full-time job, and you can now become a programmer of robots. So, it's a series of features that enable that.

One is effectively Vyde-coding your robot, where you can just describe what you're trying to do. And the robot program is built entirely on its own, and it's asking you questions and it's interactive. It's much more of kind of the quad-code type experience than legacy robot programming.

And then we also are releasing a series of features where traditionally robots, almost 99% of manufacturing robots today, people don't realize, they're just are replaying the same motions again and again. So you'll save points, kind of like the crane game at Denny's, and then the robot will replay through those points. And that's all those robots really do all day long. And so we've released vision encoder trained on billions of images, where you can really quickly train an object detection or object segmentation model to find objects and be able to pick them up. And we've made it so it's just seamless. So you can again be a, maybe just a, you are the manufacturing worker who is manually doing the job before. Now you can program a model to do that. And we're also releasing our first kind of in a series of end to end learning products. It's for a few shot learning where you can demonstrate a task to the robot. It learns from you demonstrating. And so all these things are really two axes, which we think can move that 6% number. One is making the robot so anyone can take them out of the box, making an iPhone type experience for what traditionally is more of a mainframe computer type product. And the other is making them more capable so you can just automate a lot of jobs that previously had too much variability and you could not.

2:58

SPEAKER_2

When it comes to actually vibe coding a robot, I think what makes vibe coding unique is that it's all software. And so you can tell an agent, all right, fix this, make no mistakes. And all of its context is in the actual software. How does that work when you're dealing with hardware and things in the physical world?

3:15

Evan Beard

Yeah, it's a great question. Our initial release, you have to manually program the points, but it can do so many things. So imagine you have tool calls where you just have hundreds and or thousands of example robot programs that can do almost anything. You know, like any kind of industrial task, you have examples that the LM can pull from.

So that's what we have today. You still have to manually kind of tell the robot where everything is in space. Our release later this winter, we actually scanned the environment around the robot and then it's all automatic. You can recognize all the things around the robot. So you can just say, hey, crowd these parts from the bin and put them in this vice or whatever it is. And so it's getting really real really quickly and it's amazing things. And obviously you can then close the loop agentically and you can have the robot working in the background or even overnight and refining. And we're just seeing these incredible results coming from this line of work and really exciting.

Read the other 11 minutes of this episode here.

$15 gets you this episode and 99 more. Credits never expire. After checkout, one link brings you back here to read it.

Not what you expected? Full refund within 14 days.

Thousands of transcripts fetched by people building searchable podcast archives

Fetch transcripts with code

The same credits work from a script or an AI agent. The demo key returns a sample episode in full, no card needed:

request
curl -H "x-api-key: pt_demo" \
  https://spoken.md/transcripts/1000651996090

Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.

From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.

Using your own key:

request
curl -H "x-api-key: YOUR_KEY" \
  https://spoken.md/transcripts/1000798527036