BEHIND THE ENGINEERING
REVOLUTIONIZING WEATHER DATA RECOVERY
Launching an instrument to collect information about the skies is one thing, but having it land with its equipment intact is quite another. It’s a challenge that student Yohan Hadji is poised to solve.
Written by Cassandra Kelly
FIVE YEARS AGO, Yohan Hadji set out to complete a high school project. It was a simple experiment—send a CubeSat up with a weather balloon to collect data. But as with many weather balloon missions, the device was lost.
“We knew recovery of the CubeSat would be very challenging,” Hadji said. “And I believed within three months we could develop a magic system that would make recovery easier.”
Although that didn’t happen at the time, Hadji has since forged on and dug into his research. After discovering that this issue has plagued scientists since the first weather balloons went up in the 1890s—in fact, 80 percent of radiosondes are never recovered—Hadji decided to take on the project full time.
There have been two popular methods for trying to overcome this challenge. The first is using a glider. However, gliders are often so big that they require a permit—plus, their size makes it difficult to collect data in many locations. The second method is a guided parachute, which is less reliable than the glider, but doesn’t typically require a permit.
Hadji settled on a guided parachute that looked much like a paraglider. His system launched with a balloon and once at altitude, a deployment bag released, causing it to inflate and take shape mid-air, then gracefully carry the instruments back down. He spent the bulk of his senior year in high school tinkering with the chute and the strings to keep it all tethered. By his freshman year at EPFL, the Swiss Federal Technology Institute of Lausanne, where he studied microtechnology, Hadji had a prototype that was working well enough on the ground, so it was time for a test flight.
“I took a chance and sent an email to MeteoSwiss asking if there was any way they would be willing to support me and to help me launch,” he said. “One month later, I was in the office presenting the project. I think they thought there I had some good technology, but I was still some random 17-year-old.”
Within six months, Hadji was ready for his first launch with MeteoSwiss, but when the balloon reached altitude, the system never detached. The balloon carried it nearly to Germany. By pure luck, they were still able to recover the system in perfect condition.
Hadji never understood why the system didn’t detach that day. The wind wasn’t too strong and the system didn’t have any kind of critical failure. It simply didn’t work. Fortunately, MeteoSwiss gave him a second chance, and this time, the national director of MeteoSwiss was there.
“I thought something else was going to fail, I was extremely stressed,” he said. “Then it reached altitude, the parachute deployed correctly, and somehow the control worked and the system came back. It landed five meters away from the launch site. It was almost like magic.”
Yohan Hadji, currently a technical developer at Meteomatics, shown with an early prototype in February 2024. Photo: Darrin Vanselow for PME Magazine
Gearing up
Hadji (left) worked to prepare the Meteoglider for a test flight from MeteoSwiss’s Payerne, Switzerland launch station in August 2024, while MeteoSwiss’s Gonzague Romanens (right) prepared the weather balloon. All photos: Yohan Hadji
Taking off
A Meteoglider launch from MeteoSwiss’s launch station in Payerne in August 2024. The standard weather balloon is visible above, guiding the glider up into the atmosphere at about 5 meters per second.
On the rise
For data validation purposes, a second radiosonde was also attached for the test flight. It can be seen sitting directly between the weather balloon and the Meteoglider on its upward ascent.
Back to base
After completing its mission, the glidersonde automatically returns to within 15 meters of its launch point. Here, the Meteoglider is shown shortly after landing in a test field near Lausanne, Switzerland.
Hadji demonstrated that a guided parachute system could work and he continued to work through the various challenges of developing the system for higher altitudes. For one, the glider would have to travel significantly faster at high altitudes to generate the same lift and drag forces. However, the drastic speed reduction required for safe descent after parachute deployment creates immense forces at high altitudes, which leads to chaotic and unpredictable behavior of the parachute system.
“The control has to be extremely precise and extremely well-tuned, and at the end of the day, the system didn’t work all of the time,” Hadji said. “The cost and the complexity required to recover a radiosonde, which doesn’t cost much, was not worthwhile.”
After four years, Hadji decided to go back to the drawing board. Armed with newfound understanding of aerospace, he began to pursue a glider. The small foam structure weighs just 250 grams—putting it well within the guidelines of airspace regulations—and safely houses the necessary instruments. Hadji has worked through multiple challenges, such as configuring the right set of controls and ensuring sufficient insulation to protect the batteries. Within just a year, he developed a successful model, called R2Home, which has been acquired by Meteomatics for commercialization.
“I had to believe that there was a way to take all of the software, all of the intelligence, all of the algorithms and experience that I’d developed and apply it to a smaller, simpler platform,” he shared. “So, I think that the reason that it went so fast with this new iteration is because there were a lot of things I had already discovered.”
Now, Hadji’s small glider has been tested more than 60 times and is successfully landing within 15 meters of the launch site with instruments intact and ready for reuse.
With so many test flights under his belt, Hadji isn’t stressed anymore. In addition to pursuing a master’s degree at EPFL, he is working at Meteomatics doing technical development on the project and learning the ropes of business.
“As a student, I could do whatever I wanted. I could take a lot of risks that the developers might not have been able to,” he said. “Now, they’ll take this project to the next stage. It can move so much faster and I can collaborate with many more people than before.”
Hadji is eager to keep working on the project through Meteomatics and hopes to see an industrialized version of his glider for widespread, global adoption. For now, though, he hopes to slow down, just a little.
“I sacrificed some of my childhood,” he admitted. “I spent a bit too much time working on this, but I’m definitely happy that I made this choice. Now I want to also be able to travel and enjoy my life in my 20s.”
Cassandra Kelly is a technology writer in Columbus, Ohio.
© 2025 The American Society of Mechanical Engineers. All rights reserved.