matter out of place

it's everywhere that you go

This video interview with an improbably fresh-faced Iggy Pop lives rent-free in my head.

Iggy’s explanation of the term “punk” has nothing to do with rebellion or spiky hair or leather trousers or drugs or guitars. It’s an attitude to life.

To be a punk, you just have to want to do something, even if you don’t know how; and want it badly enough to do it despite your lack of skill. And that perfectly describes the computer I want to make, and the company that I have to build to make that computer.

It’s a punk computer, if you will. Read that with the appropriate amount of eye-roll.

I started out just wanting to create a programming language. Then I realised that you can’t make a great programming language in a vacuum; you have to build it to do something. The solution was to create an operating system written in that language. But the same thing applies to the operating system; so let’s make a computer.

But here’s the thing: the same thing applies to the computer itself. To create a great computer, you need to do something great with it.

But what?

the hills are alive

On my way to Bangalore, I stopped for a week with my parents, who live in a retirement community in a “hill station” town called Dehradun, just far enough away from Delhi to have a peaceful, relatively unpolluted environment, and elevated enough to enjoy a more temperate climate.

Our Sitges buddies Walter and David visited, and we took them on a hiking trip up to the nearby hilltops of Mussoorie and Landour.

Although it was a hazy day, you could just make out the snowy peak called The Bandarpoonch (Monkey’s tail) in the lower reaches of the high Himalayas, floating on the horizon.

There’s a particular delight to hiking - you have to watch your step, so you start to notice all the tiny mountain flowers and herbs at your feet - there’s a huge variety, and our knowledgeable forest guides were only too happy to name them, describe them, and tell us stories of how they were used for various ailments in traditional Indian medicine. We even tasted a couple.

Eat your heart out, René Redzepi of Noma fame.

But I couldn’t help noticing another kind of Himalayan flora.

The roadsides along our route up and around the mountains were drowning in plastic waste. Crisp packets, paan wrappers, biscuit wrappers, empty plastic bottles, cigarette packets, bottle caps, plastic bags, drinks cans.

It’s everywhere.

And there’s more every day, courtesy of the ever increasing number of tourists drawn to the beauty of these mountains.

That serene river that runs past my parents’ house?

Look a little closer.

As heartbreaking and enraging as I found this, I also noticed that the attitude of the people that live there was basically a shrug of resignation, or straight-up indifference. It’s so endemic, and has been for so long, that it feels like people don’t even see it any more.

But I see it.

It started to get so that it was all I could see.

I didn’t just find it heartbreaking and enraging : I found it humiliating. Here’s me showing my American friends around : “Look at the charming people and scenery of my country, just please ignore the fucking crisp packets”.

Is this really who we are?

On the other hand - fuck me, right? Who the hell am I to march in with my nose in the air (having spent the last 30 years eating bonbons on my chaise longue) and declare that I’ve got a problem with this? Nobody else seems to. Better just to shut the fuck up and get used to it like everyone else, focus on the fun stuff and have a fun life, right?

I guess you already know where this is going.

“it’s my own invention”

Those who’ve had the great good fortune of working with me in the past know that there’s a part of my brain, located just beneath my medulla oblongata, that’s ever the technological egotist. (I was going to use the word optimist, but egotist is more accurate).

Like the White Knight from Alice In Wonderland, there’s no problem that it does not believe it can solve (despite ample evidence to the contrary): the Menai Bridge can be kept from rust by the simple expedient of boiling it in wine. And once this errant bit of my brain has seized upon a notion, it returns to it obsessively, nibbling and gnawing away at it despite my best attempts to silence it.

“Imagine a van with a dustbin on the back,” it said. “Imagine a van that manages a swarm of drones that pick up the litter and dump it in the dustbin.”

I described this idea to my mother, who snorted. “Congratulations! You’ve just invented the road sweeper van.”

Touché.

But a road sweeper van can’t pick the litter out of the scrub and forest that surrounds the road on a steep incline, both above it and down what the locals call the “khud”, can it? Plus, road sweeper vans exist already, and I didn’t invent them, which makes them inherently boring.

What I’m talking about is a road sweeper van with drones, which are inherently cool.

Picture it : a ramshackle Suzuki truck driving down a winding Himalayan road at 5 miles per hour, followed by a crowd of barking dogs and cursing holidaymakers leaning on their car horns as they overtake it, surrounded by a swarm of thousands of almost-invisible drones emitting an unearthly hum, darting into the surrounding undergrowth and back again.

Where the truck has been: no more plastic.

William Gibson would totally dig it, right? Because drones are cool. I mean, my friend Federico makes light shows for Celine Dion with drones, for crying out loud. In Switzerland.

If they’re good enough for Celine Dion, they’re good enough for us.

project moop

To be real, a project needs a codename.

“moop” is an acronym: it stands for matter out of place. It’s a Burning Man term for litter (yes, I went in 1998, and it was way better then).

Project moop deliberately does not concern itself with the goal of cleaning up the entire Indian subcontinent using a fleet of millions of vehicles and billions of drones. Like the notion of creating a media server company, that’s clearly a ridiculously impossible dream and you shouldn’t even talk about something that crazy, let alone plan to actually do it.

Instead, project moop is about building three super easy demos.

demo 1: the sucker

The Sucker is a 3D-printed soft-robotics suction-based gripper based on an Octopus tentacle suction cup, about the radius of a fingertip, with enough suction to pick up a range of small plastic litter items.

By default, an actuator squeezes the inner gas chamber closed; when current is applied, it opens, creating suction. Pretty simple, and it works for the octopus, so it should work for us. Add a touch sensor that triggers the actuator, and the sucker will grab onto anything it touches.

Instead of a tentacle, however, we’re going to use a rubber band to mount the sucker inside a cylindrical barrel housing (think: sawn-off sharpie), with some kind of spring-loaded mechanism inside that can fire it, on command, at a target from some fixed distance. The sucker hits the target, the touch sensor triggers the actuator, the sucker grabs the target, and the rubber band then contracts and returns the sucker to the housing. Then we need some kind of electromagnetic mechanism to re-arm the spring system and we’re ready to repeat.

Easy!

demo 2: mount the sucker on a drone

Since you’re going to eventually needs lots of drones, we find the simplest and cheapest drone we can get away with. Turns out there’s a ton of open source designs out there for tiny little DIY handheld drones like this:

For demo 2, we mount the sucker on a drone, and fly it (under computer control) from a home base out over a target at a known location, high enough so the prop wash doesn’t push the target away; fire the sucker downwards to grab the target; and then return to the home location and turn off the actuator, dropping the target item into a bin.

Easy!

demo 3: mount two drones on a vehicle

Next, we get hold of a camera and some image-classifier software running on a Raspberry Pi computer, and train it to recognise our plastic menagerie. Stick the camera and the Pi onto a radio-controlled dump truck.

The demo is simple: we drive the truck (at a low fixed speed) past a range of randomly placed litter items; the vision system identifies them and locates them; the controller assigns each drone to a different target location; and sends them out repeatedly until all the targets have been collected. If we can make it work with two drones, we can make it work with a thousand; and if we can make it work with a toy dump truck, we can make it work with a real one.

Easy!

build it, and they will come

So here’s the thing: I don’t know what happens next. What I do know is that if we get these three super easy demos working and put them out there, the right people will show up to take the next step.

How do I know this?

Because I pitched it at a startup networking meeting this morning in Bangalore.

Turns out Iggy’s not the only punk out there.

Next Post.