As a kid, Jason Huggins was “mechanically minded,”­­ and eager to build things. When his high school English teacher assigned the class to summarize the plot of Romeo and Juliet, Huggins completed the project in the form of a pop-up book.

After college, Huggins took a software job. But he felt that something was missing. It wasn’t “tangible, hands-on, tactile–something I can put on the table and say ‘hey, I made that.'” He had always been interested in robotics, and so he taught himself the electronics and mechanical skills needed to create the machines, including 3D printing, electronics design, milling, and laser cutting. He’s since worked at Google, started a software company, and even helped fix HealthCare.gov.

Now, Huggins runs a company called Tapster that makes robots to test if mobile devices are working by mimicking a human finger that taps on the screen. He describes his work as making robots that “solve problems for companies.” But on the side, he is still committed to “robotic art projects.” Huggins calls himself “a robotics entrepreneur in support of being a robotics artist.”

Here are five lessons Huggins has learned from his many years pursuing “robot art.”

1) Work on projects that don’t have a purpose

At a software conference five years ago, Huggins showed off a robot with arms he made for fun–and to help play Angry Birds. It had a robotic arm that you could send a signal to from a computer, and would move to any point on an iPhone screen. The arm would move down to the screen surface and click or drag objects–in order to pull and release the slingshot for the game. The robot wasn’t particularly high-functioning, unable to progress beyond level 1. But, as it turns out, that didn’t matter. Observers noticed that the machine could serve another purpose: Testing apps on mobile touchscreens.

When you put something out there, Huggins said, “people come out of the woodwork” and help conceive of new functions for the creation. Huggins said he suspects that many inventions happen like this, when “someone was just kind of fumbling around, and the practical use came out later.”

Huggins said he didn’t realize that people who were working at big phone or car companies were already using robots to do something similar in their secret quality control labs. Before he started working with Mercedes-Benz, it was testing a remote parking app, to let people “get out of the car, tap the button on the phone, it parks the car.” They needed a robot to handle the phone the way a human would, and were looking for an inexpensive robotic arm to test it on a real phone. “They were about to buy a Lego Mindstorms kit, like a box of Lego, and dedicate two people on the team to spend a couple months, [saying] here’s a budget to buy some Legos and figure it out,” Huggins said.

SEE: Photos: See how ‘Tapster’ robots imitate human touch on mobile devices

2) Adopt a software developer’s mindset

Part of the success of the Angry Birds robot was the fact that it was related to software development. A big trend, Huggins said, is that all companies are now becoming software companies. Thanks to Tesla, auto companies are increasingly looking to develop systems for over-the-air updates. That means more touch screens and panels–“a lot of companies are having to figure out how to become software shops and replacing things with touch panels and things like that,” he said.

3) Look beyond industrial or educational robots

When Huggins was in the fourth grade, one of his favorite toys was a little robotic arm called Robotron from Radio Shack. As he grew older, Huggins wanted to work in robotics, but didn’t see how it could turn into a day job. One of the problems, he said, is that there are these big industrial robots, like the ones in car factories, or, on the flip side, the educational type of robots.

If kids are inspired to play with robots, but only have a chance to build them in a factory setting, “it’s a missed opportunity to go solve all of those problems people complain about,” said Huggins, “like not having robots do their dishes or mow your lawn.”

Part of the reason, he said, is because there’s a lot of money to be made in factories where robots are mass producing items. And the household setting doesn’t require the same kind of production. “No one’s making a thousand peanut butter and jelly sandwiches every morning,” he said.

Still, with the price of robotic components and motors coming down, making “little robots to do little things” isn’t quite as cost-prohibitive anymore, Huggins said. As the costs of things like 3D printers and circuit boards drop, tinkering around and making robots for fun has become more accessible for those on smaller budgets. And with mobile phone components becoming “smaller and more ubiquitous and cheaper, there’s this perfect wave where, if you want to make something mechanical that moves, the global supply chain is making that easier and easier to do,” Huggins said.

“It’s easier to justify putting in a little bit of effort to make a little robot do something,” he added. “Also, you don’t necessarily have the market for it. You can just make it because you wanted to make it and it didn’t cost $100,000 to do it.”

4. Create multidisciplinary teams

In the past, tasks were likely to be siloed. You had software developers in one corner, then testers in another, then operations, said Huggins. But more recently, innovation is happening in collaborative spaces. Processes are becoming interdisciplinary, said Huggins, like in agile software development.

SEE: The next big job in tech: Robot programmer

He said he thinks that universities should integrate this model, especially when it comes to robotics. For example, mechatronics is a major a school can offer that integrates the software, electronics, and mechanics involved in robotics. But there aren’t a lot of mechantronic programs around, he said. There are mechanical engineer programs that occasionally do robotics, or electrical engineering departments that integrate software. “I think they all suffer from the fact that they’re siloed in their one major department,” he said.

5. Reimagine art

Huggins said he thinks that technological advances will bring about changes in what–and how–the public views art. In modern art museums, for instance, he predicts that the static way art is displayed will change. Because of the lower cost of electronic components, he thinks that art will begin to be more kinetic, with LEDs and motors. Huggins says this is all part of an art movement, “electronic modernism.”

“In 100 years, I think there’s going to be a lot of art that you have to plug in,” he said. “I’m really interested in exploring that space. Everything I do from a career or business aspect is really kind of to fund pushing those art questions.”

SEE: Video: The ‘Queen of Shitty Robots’ explains why creating without a goal can be empowering

Huggins has been working on one project for 15 years. “I was making a 3D simulation of waves on my computer screen, and I thought it was interesting that it was trapped behind a 2D piece of glass,” he said. “It’s a projection that basically wanted to be 3D but couldn’t because it was in a flat computer screen. Next to my desk I had this pin art display–an array of plastic or metal pins.” Huggins had an idea: What if he could motorize the pins to create a moving 3D display, like the one he was animating on his screen?

Huggins calls it the “reverse tron,” where you can remove the 3D animation from the computer screen and bring it into the real environment. Over the years, he said, he’s been trying to hone the skills to complete the project. “Everything else I’ve been doing, including even the robots I’ve made, have basically been kind of smaller spinoffs from that bigger project that I’m still pursuing,” he said.

“Now that I’ve stumbled into that idea behind it, of taking things that are trapped inside a computer screen and bringing them to life in some kind of 3D form, I realize that that could be kind of a theme,” he said. “I can’t die until I complete that project.”

Also see…