There’s an interesting development in amateur ballooning: using so-called superpressure balloons, which float high in the atmosphere indefinitely rather than simply going up and up and then popping like a normal weather balloon. Superpressure balloons can last for months and travel long distances, potentially circumnavigating the globe, all the while reporting their position.

You might imagine that an undertaking like this would be immensely difficult and cost thousands of dollars. In fact, you can build and launch such a balloon for about the cost of a fancy dinner out. You just have to think small! That’s why amateur balloonists call them pico balloons.

The payload of a pico balloon is so light (between 12 to 30 grams) that you can use a large Mylar party balloon filled with helium to lift it. They’re also inexpensive; that’s important because you won’t get your payload back. And because such diminutive payloads don’t pose a danger to aircraft, they aren’t subject to the many rules and restrictions on free-floating balloons that carry more mass.

The essential advances that made pico ballooning possible were figuring out how to track a balloon no matter where in the world it might be and how to power such tiny payloads. A lot of folks worked on these challenges and came up with good solutions that aren’t hard or expensive to reproduce.

What is WSPR?

Amazingly, the global tracking of the balloon’s telemetry is done without satellites. Instead, pico balloonists take advantage of an amateur-radio network called WSPR (Weak Signal Propagation Reporter), a protocol developed by a rather famous ham-radio enthusiast—Joseph Hooton Taylor Jr., one of the two scientists awarded the 1993 Nobel Prize in Physics for discovering binary pulsars.

A Raspberry Pi Pico microcontroller [top left] is soldered directly to a daughterboard consisting of a high-frequency transmitter and a GPS module [bottom left], which are all powered by solar panels [right].James Provost

WSPR was designed to monitor signal-propagation conditions for different radio bands—useful information if you’re a ham trying to make distant contacts. WSPR can also record low-power balloon-telemetry signals. WSPR is very low bandwidth—less than 10 bits per minute—but it does the job. A worldwide network of radio amateurs receives these WSPR signals and reports them publicly over the internet, which gives picoballoonists a way to track their flights. You need at least a general-class ham-radio license to launch a pico balloon, as one is required to transmit on the bands used for long-distance telemetry.

The pico balloon payload I chose to build is based on the aptly named US $4 Raspberry Pi Pico board, with a solder-on daughterboard that contains a GPS receiver and transmitter. The folks who developed this daughterboard and associated software (to create what they call the Jetpack WSPR Tracker) have done a fantastic job of making their work easy to reproduce.

You could, in principle, power the Jetpack tracker with batteries, but in practice it would be impossible to keep them warm in the stratosphere, where average temperatures can be as low as –51 °C. Instead, the tracker runs off two lightweight solar modules. At night, it gracefully powers down. When the sun rises high enough in the morning, the tracker powers up and starts transmitting again.

My first pico balloon made it only halfway across the Atlantic before going silent.

I had five Jetpack boards custom-manufactured in China for just $39. The cost nearly doubled after adding shipping and tariff charges. Still that’s really cheap, even when you add the cost of the Raspberry Pi ($4), the party balloon ($10 for two), the helium ($10 at my local supermarket), and the two solar modules ($7 each).

The biggest sticking point I had with the Jetpack design was the liberties it takes with spurious emissions from its transmitter. Federal Communications Commission (FCC) regulations call for spurious emissions to be at least 43 decibels below the power of the transmitted signal. But my transmitter had strong unwanted emissions at odd harmonics of the fundamental frequency. (That’s because the transmitter is a Si5351A temperature-controlled oscillator, which outputs a square wave, not a sinusoid.) Taking measurements, I could see that the third harmonic at 42 megahertz was only 25 dB quieter than the 14-MHz fundamental of my WSPR signal’s frequency.

As of press time, the WSPR network had tracked my balloon from the Eastern United States to the Mediterranean coast. James Provost

In practical terms, this shouldn’t create any noticeable interference, given that this transmitter puts out milliwatts at most and floats miles away from the nearest receiver. Still, I wanted to be fully compliant with FCC regulations, so I added traps to the antenna—simple circuit elements that hams use to allow a single antenna to work on multiple bands by altering how the antenna resonates at different frequencies. Each trap was made of a small inductor (four 5-millimeter-diameter loops of No. 32 magnet wire) in parallel with a 220-picofarad capacitor. I tuned them with the help of a NanoVNA signal analyzer by stretching the loops apart slightly. I attached the traps directly to the tracker board, so that they quashed the spurious 42-MHz emissions at the source. That worked well and added only 0.3 grams of weight.

With my payload complete, I partially filled my balloon with helium. You want the balloon to hold just a little more gas than it takes to lift the payload off the ground. This will give the helium room to expand as the balloon climbs to its final altitude.

My first pico balloon, launched from a park near my home in North Carolina, made it only halfway across the Atlantic before going silent. My second went up and was never heard from again. The third was indeed the charm. It crossed the Iberian Peninsula and at the time of this writing is somewhere over the Mediterranean at an altitude of nearly 12 kilometers. With any luck, it might go on to orbit the planet.

I’m a little puzzled about the balloons’ telemetry messages received on the WSPR network, as they have been few and far between. My best guess is that power from the horizontal solar panels I’m using is marginal, with the winter sun being so low in the sky. That’s something I should have thought about before launching the first balloon just 24 hours after the winter solstice!

This article appears in the February 2026 print issue as “Long-Duration Amateur Ballooning.”

Assistive technology is expensive, and many people with disabilities live on fixed incomes. Disabled assistive tech users also must contend with equipment that was often designed without any capacity to be repaired or modified. But assistive tech users ultimately need the functionality they need—a wheelchair that isn’t constantly needing to be charged, perhaps, or a hearing aid that doesn’t amplify all background noise equally. Assistive tech “makers,“ who can hack and modify existing assistive tech, have always been in high demand.

Therese Willkomm, emeritus professor of occupational therapy at the University of New Hampshire, has written three books cataloging her more than 2,000 assistive technology hacks. Wilkomm says she aims to keep her assistive tech hacks costing less than five dollars.

She’s come to be known internationally as the “MacGyver of Assistive Technology” and has presented more than 600 workshops and assistive tech maker days across 42 states and 14 countries.

IEEE Spectrum sat down with Willkomm ahead of her latest assistive tech Maker Day workshop, on Saturday, 31 Jan., at the Assistive Technology Industry Association (ATIA) conference in Orlando. Over the course of the conversation, she discussed the evolution of assistive technology over 40 years, the urgent need for affordable communication devices, and why the DIY movement matters now more than ever.

IEEE Spectrum: What got you started in assistive technology?

Therese Wilkomm: I grew up in Wisconsin where my father had a machine shop and worked on dairy and hog farms. At age ten, I started building and making things. A cousin was in a farm accident and needed modifications to his tractor, which introduced me to welding. In college, I enrolled in vocational rehabilitation and learned about rehab engineering—assistive technology wasn’t coined until 1988 with the Technology-Related Assistance Act. In 1979, Gregg Vanderheiden came to the University of Wisconsin-Stout and demonstrated creative things with garage door openers and communication devices. I thought, wow, this would be an awesome career path—designing and fabricating devices and worksite adaptations for people with disabilities to go back to work and live independently. I haven’t looked back.

You’ve created over 2,000 assistive technology solutions. What’s your most memorable one?

Wilkomm: A device for castrating pigs with one hand. We figured out a way to design a device that fit on the end of the hog crate that was foot-operated to hold the hind legs of the pig back so the procedure could be done with one hand.

Assistive Technology’s Changing Landscape

How has assistive technology evolved over the decades?

Wilkomm: In the 1980s, we fabricated devices from wood and early electronics. I became a [Rehabilitation Engineering and Assistive Technology Society of North America, a.k.a. RESNA] member in 1985. The 1988 Technology-Related Assistance Act was transformational—all fifty states finally got funding to support assistive technology and needs in rural areas. Back in the ‘80s, we were soldering and making battery interrupters and momentary switches for toys, radios, and music. Gregg was doing some things with communication. There were Prentke Romich communication devices. Those were some of the first electronic assistive technologies.

The early 1990s was all about mobile rehab engineering. Senator Bob Dole gave me a $50,000 grant to fund my first mobile unit. That mobile unit had all my welding equipment, all my fabrication equipment, and I could drive farm to farm, set up outside right in front of the tractor, and fabricate whatever needed to be fabricated. Then around 1997, there were cuts in the school systems. Mobile units became really expensive to operate. We started to look at more efficient ways of providing assistive technology services. With the Tech Act, we had demonstration sites where people would come and try out different devices. But people had to get in a car, drive to a center, get out, find parking, come into the building—a lot of time was being lost.

In the 2000s, more challenges with decreased funding. I discovered that with a Honda Accord and those crates you get from Staples, you could have your whole mobile unit in the trunk of your car because of advances in materials. We could make battery interrupters and momentary switches without ever having to solder. We can make switches in 28 seconds, battery interrupters in 18 seconds. When COVID happened, we had to pivot—do more virtual, ship stuff out to people. We were able to serve more individuals during COVID than prior to COVID because nobody had to travel.

How do you keep costs under five dollars?

Wilkomm: I aim for five dollars or less. I get tons of corrugated plastic donated for free, so we spend no money on that. Then there’s Scapa Tape—a very aggressive double-sided foam tape that costs five cents a foot. If you fabricate something, and it doesn’t work out, and you have to reposition, you’re out a nickel’s worth of material. Buying Velcro in bulk helps too. Then Instamorph—it is non-toxic, biodegradable. You can reheat it, reform it, in five minutes or less up to six times. I’ve created about 132 different devices just using Instamorph. A lot of things I make out of Instamorph don’t necessarily work. I have a bucket and I reuse that Instamorph. We can get six, seven devices out of reusable Instamorph. That’s how we keep it under five dollars.

What key legislation impacts assistive technology?

Wilkomm: Definitely the Technology-Related Assistance Act. In the school system, however, it only says “did you consider assistive technology?” So that legislation really needs to be beefed up. The third piece of legislation I worked on was the AgrAbility legislation to fund assistive technology consultations and technical assistance for farmers and ranchers. The latest Technology-Related Assistance Act was reauthorized in 2022. Not a whole lot of changes—it’s still assistive technology device demonstrations and loans, device reuse, training, technical assistance, information and awareness. The other thing is NIDILRR—National Institute on Independent Living and Rehabilitation Research, funded under [the U.S. Department of Health and Human Services, a.k.a. HHS]. Funding the rehab engineering centers was pretty significant in advancing the field because these were huge, multimillion-dollar centers dedicated to core areas like communication and employment. Now there’s a new one out on artificial intelligence.

A Vision for a Better Assistive Tech Future

Over more than 2,000 hacks to improve usability of assistive technologies, veteran DIY maker Therese Wilkomm has earned the moniker “the MacGyver of assistive tech.” Therese Willkomm

What deserves more focus in your field?

Wilkomm: The supply-and-demand problem. It all comes down to time and money. We have an elderly population that continues to grow, and a disability population that continues to grow—high demand, high need for assistive technology, yet the resources available to meet that need are limited. A few years back, the Christopher & Dana Reeve Foundation had a competition. I submitted a proposal similar to the Blue Apron approach. People don’t have supplies at their house. They can’t buy two inches of tape—they have to buy a whole roll. They can’t buy one foot of corrugated plastic—they’ve got to buy an 18-by-24 sheet or wait till it gets donated.

With my third book, I created solutions with QR codes showing videos on how to make them. I used Christopher Reeve Foundation funding to purchase supplies. With Blue Apron, somebody wants to make dinner and a box arrives with a chicken breast, potato, vegetables, and recipe. I thought, what if we could apply that to assistive technology? Somebody needs something, there’s a solution out there, but they don’t have the money or the time—how can we quickly put it in a box and send it to them? People who attended my workshops didn’t have to spend money on materials or waste time at the store. They’d watch the video and assemble it.

But then there were people who said, “I do not have even five minutes in the school day to stop what I’m doing to make something.” So we found volunteers who said, “Hey, I can make slant boards. I can make switches. I can adapt toys.” You have people who want to build stuff and people who need stuff. If you can deal with the time and money issue, anything’s possible to serve more people and provide more devices.

What’s your biggest vision for the future?

Wilkomm: I’m very passionate about communication. December 15th was the passage in 1791 of our First Amendment, freedom of speech. Yet people with communication impairments are denied their basic right of freedom of speech because they don’t have an affordable communication device, or it takes too long to program or learn. I just wish we could get better at designing and fabricating affordable communication devices, so everybody is awarded their First Amendment right. It shouldn’t be something that’s nice to have—it’s something that’s needed to have. When you lose your leg, you’re fitted with a prosthetic device, and insurance covers that. Insurance should also cover communication devices and all the support services needed. With voice recognition and computer-generated voices, there are tremendous opportunities in assistive technology for communication impairments that need to be addressed.

What should IEEE Spectrum readers take away from this conversation?

Wilkomm: There’s tremendous need for this skill set—working in conjunction with AI and material sciences and the field of assistive technology and rehab engineering. I’d like people to look at opportunities to volunteer their time and also to pursue careers in the field of specialized rehab engineering.

How are DIY approaches evolving with new technologies?

Wilkomm: What we’re seeing at maker fairs is more people doing 3D printing, switch-access controls, and these five-minute approaches. There has to be a healthy balance between what we can do with or without electronics. If we need something programmed with electronics, absolutely—but is there a faster way?

The other thing that’s interesting is skill development. You used to have to go to college for four, six, eight years. With YouTube, you can learn so much on the internet. You can develop skills in things you never thought were possible without a four-year degree. There’s basic electronic stuff you can absolutely learn without taking a course. I think we’re going to have more people out there doing hacks, asking “What if I change it this way?” We don’t need to have a switch.

We need to look at the person’s body and how that body interacts with the electronic device interface so it requires minimal effort—whether it be eye control or motion control. Having devices that predict what you’re going to want next, that are constantly listening, knowing the way you talk. I love the fact that AI looks at all my emails and creates this whole thing like “here’s how I’d respond.” I’m like, yeah, that’s exactly it. I just hit select and I don’t have to type it all out. It speeds up communication. We’re living in exciting times right now.