I grew up watching Star Wars, Star Trek, Buck Rogers and Knight Rider, waiting for the day when science fiction would become reality.

It wasn't only the spaceships, distant planets or futuristic cities that captured my imagination. I was fascinated by the machines that were such an integral part of those stories. More specifically, the robots. R2-D2 and C-3PO bickered and worried like old companions. Twiki was mischievous and irreverent. KITT was intelligent, dignified, loyal, and occasionally more human than the people around him. Even the ship's computer aboard the Enterprise had a kind of reassuring presence.

These machines didn't just perform tasks. They listened. They responded. They remembered. They had voices, attitudes, and relationships with the humans around them.

I assumed that was where technology was heading, that one day we'd speak naturally with intelligent machines, and they'd know us, help us, and occasionally surprise us. Didn't we all think that?

Decades passed, and the future arrived in pieces. Computers shrank down to fit in our pockets. The internet connected nearly every corner of the world. Phones became cameras, maps, libraries and gateways to almost everything we knew. Artificial intelligence learned to recognize voices, interpret images, and carry on remarkably natural conversations.

And yet the robot companion I'd imagined as a child never quite showed up, and the few now being promised remain prohibitively expensive for most people.

So why not build one myself?

A Robot Out of Time

Robie Sr. came from RadioShack and belonged unmistakably to the optimistic world of the mid-1980s: a smiling plastic face, flashing lights, motorized wheels, and the confident look of a machine that should have been capable of extraordinary things.

In reality, he was an expensive toy. He could roll around, beep, light up his eyes and mouth, and play cassettes. He couldn't see you, understand you, or decide anything. He had no memory and no personality beyond whatever a child projected onto him. Like most "robots" of that era, Robie was an elaborate remote-controlled gadget. The intelligence lived entirely in the imagination of whoever was holding the controller.

But the promise was there.

Nearly forty years later, looking at one of these old robots, I found myself wondering whether time had finally caught up with him. The technologies that existed only in science fiction when Robie was built, speech understanding, object recognition and natural conversation, were now part of modern technology. Robie himself hadn't changed at all. Everything around him had.

I started wondering what would happen if I brought those two timelines together: the body of a robot from my childhood, combined with the intelligence of the present. Could Robie finally become the character his appearance had always suggested he was?

Beginning Without a Map

I didn't start with a grand plan. There was no blueprint for exactly what Robie would become. The project grew one idea at a time, each one leading almost inevitably to the next.

At first I just wanted to bring him back to life: get his eyes and mouth reacting, get him talking. Once he could speak, I wanted him to listen. Once he could listen, I wanted him to understand. Once he could understand, I wanted him to remember things and bring them back up in conversation. Once he could see, I started wondering if he could interpret what was in front of him.

Before long, Robie could respond to natural speech, answer questions, describe what he saw, and move on command. He could take photos, inspect objects, remember details, and keep watch over a room. None of that mattered to me as a checklist of features. What interested me was what happened when it all worked together, whether his voice, movement, memory and responses could create the impression that somebody was home.

The Trouble With Time

Time behaves strangely when you build something this way.

Months disappear into problems that sound almost comically small once you describe them. An entire evening can vanish into a wiring issue, an unexpected sound, or a robot that suddenly refuses to listen. Something that worked yesterday can behave inexplicably today.

There were broken connections and failed parts along the way. A sound card had to be replaced. A speaker got punctured. A camera was damaged and swapped out. Motors occasionally did things I hadn't asked for and didn't especially want. Software fixes aimed at one problem sometimes created two more. Robie would interrupt me mid-sentence, or ignore me entirely.

The project demanded patience, and just as often, the willingness to start over. More than once I've had to admit a promising idea wasn't working, roll it back, and return to something stable. Building Robie taught me that progress isn't always about adding more. Sometimes it's about not breaking what already worked.

But every so often, something clicked. He looked through his camera and described what was in front of him. He responded when I asked him to move. He remembered something I'd told him days earlier. He made a remark that fit the moment so well it felt less like a scripted output and more like the beginning of a personality.

Those moments are what keep the project going.

Creating a Character

The machines I loved growing up were memorable because they had character. KITT had logic and restraint. C-3PO was a comedic pessimist. R2-D2 was brave and a little reckless. Twiki was mischievous. Their personalities came through in voice, timing, movement, and the way they responded to the people around them.

Robie needed that same quality. A machine can answer a complicated question and still feel completely lifeless, plenty of processing power, no sense of connection. I wasn't trying to build a chatbot in a plastic shell.

This is where my background ended up mattering more than I expected. I graduated from the University of Toronto with a double major in Psychology and Kinesiology. I didn't set out to apply either discipline to a robotics project, but much of what makes a character feel present sits somewhere between the two. Psychology helped me understand how we read intention, attention and personality into behaviour. Kinesiology gave me an appreciation for how movement itself communicates, how timing, posture and even a slight hesitation can make an action appear deliberate rather than mechanical.

A pause before an answer, a reference to something said days earlier or a movement at just the right moment can completely change how we perceive a machine. Personality doesn't come from any single feature. It emerges in the spaces between them, where memory, intelligence and a distinct persona begin to coalesce into something that feels remarkably alive.

None of this made the engineering any easier. I was self-taught on the electronics side, and my coding experience amounted to a couple of high-school computer science courses that had long since become rusty. Every obstacle forced me to learn something new, and every small success expanded my idea of what I might still be capable of building.

When He Finally Talked Back

The first time Robie responded naturally to something I said, unprompted by any remote control, the gap between the 1980s and the present narrowed considerably. His blinking eyes and moulded smile were still exactly as they'd always been. But I spoke, he listened, and something came back unscripted in that moment.

The child who watched Star Wars, Star Trek, Buck Rogers and Knight Rider had been waiting a long time for that exchange. The future didn't arrive the way television promised. There are no passenger flights to distant planets, and intelligent robot companions haven't become standard household fixtures. But some of those childhood predictions turned out to be within reach of one person with a soldering iron, a hobbyist's budget and a willingness to spend unreasonable amounts of time solving very small problems. The difference is that I didn't need to wait for a company to package it up and sell it to me.

That's probably the most useful way to think about the period we're in. Ideas that used to belong exclusively to corporations, research labs and film studios are increasingly accessible to individuals willing to put in the hours. The gap between imagining something and actually attempting it has gotten a lot smaller.

An Unfinished Future

Robie is still very much a work in progress, and I've started to suspect that's simply his permanent state. Every new ability exposes another limitation. Every solved problem raises a more interesting question. I used to think the goal was to "finish" him. I'm no longer sure that's the right frame for a robot designed to keep learning and changing.

I'd like Robie to become more autonomous, not simply waiting for instructions, but occasionally choosing to look around, ask a question or make an observation on his own. I'd like the line between operating a machine and simply spending time with him to become increasingly difficult to identify. At the same time, I want to preserve the qualities that make him unmistakably Robie: the same smiling plastic face and the same 1986 body, now animated by ideas that would have sounded like science fiction in the year he was built.

Robie spent forty years looking like a robot from the future.

Only now is he beginning to become one.