Yelp’s Dev Team Beats NASA, Sends First 3D Printer to the Edge of Space

This slideshow requires JavaScript.

Gathering in a gravel lot alongside the train tracks and Delta waterway of Antioch, Calif., 45 miles east of San Francisco, a scrappy team of web engineers from Yelp spent the early part of this past Sunday morning organizing supplies to attempt to trump NASA’s latest ambitious goal: bring a functioning 3D printer to space, or at least much closer to it. By the end of the day, they’d reach their personal target, sending their package into the stratosphere somewhere between 90,000 to 100,000 feet above the Earth’s surface, not quite the Kármán line, but past where the curvature of the planet is easily seen and the sky above is shrouded in murky darkness.

NASA and the space science community have grand plans for extraterrestrial 3D printing—lunar bases extruded from moon dust, astronaut tools laser-sintered from regolith, even space pizza printed from powder. Despite such futuristic goals, the agency has yet to attempt any 3D printing technology in the harsh extremes beyond Earth’s atmosphere. It’s an endeavor they’re pursuing actively, though; after successfully 3D printing in near-Zero-G conditions on the “vomit comet” trial flights, NASA recently announced plans to test a specially designed FDM-style 3D printer on the ISS in 2014.

It would be the first 3D printer in space, unless someone were to get one there beforehand—which is what Yelpers Yoni De Beule, Cameron Paul, Mason Glidden, and Ben Asher, and ex-Yelper John Boiles set out to accomplish this past weekend in a project sponsored by their company. Their approach to get the machine high above the Earth employs the popular DIY upper-atmosphere weather balloon technique with a few fancy additions. A 1600-gram helium-filled weather balloon would lift the payload containing a diminutive, sub-five pound, Li-Po battery-powered Printrbot Simple. Tracking would occur by a custom Ardutracker: an Arduino Uno controlling a GPS and ham radio transmitter, allowing for long-distance monitoring using APRS digital communication through an affixed antenna (something each person stumbled over at least once over the course of the morning). A GoPro Hero 2 camera would capture the dramatic footage, a golden radar reflector would help alert any nearby aircraft of the balloon’s presence, and a small parachute would bring the contents, assembled into a large styrofoam container with the sides cut out, safely to the ground once the balloon popped.

While finishing the electronics assembly on a card table next to the open tailgate of Boiles’ compact SUV, the team explains to me that the 3D printer would begin to print 40 minutes after liftoff. Brook Drumm, Printrbot’s creator who was on hand for the liftoff, raises his eyebrows upon hearing the temperatures to which the printer will be subjected: approximately 60 degrees below zero Fahrenheit. He mentions he’s unsure that it will extrude at that point, but remains excited that his tiny, $299 printer is making the trip with its battery capabilities: “In a situation like this, where you going to find a power cord?”

The final pieces are put in place as one of their colleagues in a personal aircraft begins to circle overhead, in hopes to get aerial photos of the liftoff. A full tank of helium makes the balloon, inflated to about 7 feet in diameter, particularly unwieldy to hold in place. When it rises into lower pressures of the high altitude, it will have increased in diameter to almost 30 feet before bursting.

Finally, liftoff occurs, and the team cheers as the balloon quickly float upwards, directly into the overhead mid-day sun. It disappears after just a minute. The group jumps on their computers and cell phones to track the balloon’s movement, transmitted to a live map embedded on their website, darwinaerospace.com. A mad dash follows, with cars pulling out of the lot in convoy formation, chasing after the blinking dot that moves slowly northeast over the Delta’s inlets and waterways.

The team has done this twice before, although never with a 3D printer. The first time, an attempt to create the first Yelp check-in “from space,” had just one major hangup: the crew was unable to find the payload for four months after it rushed up and away from the liftoff point. They finally tracked it down to the desert north of Reno, using an autonomous drone and ATVs to pinpoint its location.  The lessons learned led another, simpler launch, and to Burrito Bomber, the fixed-wing autonomous aerial food delivery system that gave them notoriety not just at Yelp but online as well.

I follow the team in my car as we wind down country roads and atop levees through small agricultural towns. A few U-turns are made as bridges need to be crossed to stay on the right path. Finally, the group pulls over and says that the balloon is moving slowly laterally, so they’re stopping for food.

I sit down at a Mexican restaurant with Boiles, Glidden and Paul hunched over their phones calling out stats between bites. “42,000 feet” “20mph and steady” “APRS looks overloaded.” The group is smiling, very pleased with how smooth the day is moving. We start sharing stories of our backgrounds and discussing wifi-enabled cameras when Glidden goes quiet. He’s noticed that the balloon’s tracker has stopped transmitting.

Perhaps it’s the firmware, Boiles wonders, or maybe the coldness of the higher altitude. No one mentions losing the balloon into the desert again, but the group gets a bit more serious. The check is requested, and everyone packs back up in the car to keep moving towards the predicted landing zone. The weather balloon community has developed tools that accurately show the forecasted path of a balloon based on wind speed and direction at various elevations. Ours is supposed to move northeast for a while, then push southeast as it reaches the wind currents higher up, and then back to the northeast as it descends, into a location a bit further north of us. Still, if the tracking signal doesn’t come back on, it will be nearly impossible to find the package when it lands.

We continue through the small side roads and towns alongside the levees when we suddenly enter the Sacramento city limits. What had felt like a 15 minute race through the farms had actually taken us deep into central California. I follow behind as we exit into a gas station. Glidden walks over to tell me we’ll wait there in hopes of getting a signal, so it might be a bit. A moment later, he runs back over and says the tracker just pinged them—and it’s descending just a couple miles south of where we are.

The newly recovered path leads us into a quiet residential neighborhood on the southern side of Sacramento. As Boiles drives, Glidden and Paul watch the payload plot itself on their map. On one street they pull over and stare straight overhead, telling me the payload should be about 7000′ directly above us. We zig and zag a few streets over when Glidden jumps out again and points behind us, shouting “I see it!”

Indeed, there it is, a styrofoam box dangling below a small parachute, slowly floating towards the middle of the neighborhood a block or two away. The incredible accuracy of our position was not lost on me, especially with the team’s previous four-month recovery journey. They excitedly explain that it’s not common to watch the payload coming down.

All that is left at that point is to find the package, which may have landed in a street, a tree, or the most challenging, someone’s back yard. We split up and slowly patrol each block. After 45 minutes, Boiles and Paul decide to knock on some doors. Incredibly, the first one they try, a couple who had just pulled into the driveway, find the package in their back yard. Smiles and handshakes are exchanged as we see the package close-up for the first time since liftoff.

We bring everything back to the car and investigate. The styrofoam frame survived impact better than expected. Sadly, there is no print on the platform, but the GoPro video shows that the timer circuitry worked as planned—the machine activated and the printer head moved about in an attempt to start the design. We look at the printer and see that the filament is jammed in the extruder, indicating that, as suspected, the stratosphere’s low temperatures kept it from working correctly.

The team finishes the day in high spirits. “We had a very clean launch and a record speed recovery,” De Beules follows up to me in an email. “On top of that we got some awesome footage and photos from both the launch (including aerial!), the flight (although some bits are a little foggy) and the recovery. Although the printer did not print, it did go to at least 90k feet. And most importantly, we all had a really good time!”

And, they’ve just received a thumbs up to send another attempt at the end of August, with modifications that will try to increase the potential for a successful stratospheric print. We’ll be watching. Perhaps NASA will be too.