Hungarian Scientists win NASA Computer Model Challenge


Axel Ország-Krisz (left) and Richárd Ádám Vécsey.

Freelance data scientists and deep learning developers Axel Ország-Krisz and Richárd Ádám Vécsey, longtime friends and collaborators, won the top spot and USD 25,000 in the Aftershock: NASA Shock Propagation Challenge in mid-July. The Budapest Business Journal spoke with them about their solution and plans.

BBJ: How did you become interested in space-related research?

Axel Ország-Krisz & Richárd Ádám Vécsey: Every child wants to be an astronaut. In 2020, we had to decide about our free time due to COVID-19 lockdowns. We decided to move further and find challenges from our couch. This sounds simple, but that was not an easy decision. We are open and like to learn new things. We began running hackathons, and we ran into the Space Apps Challenge 2020 organized by NASA, an annual challenge for developers and space enthusiasts worldwide. We submitted a project that took us to the global finalists. This acknowledgment gave us a huge nudge. We had a lot of hackathon experiences, but NASA is a “different world.” It organized the event with absolute professionalism and enthusiasm and replied to our messages even if we sent them at quite strange times. In our defense, managing time zone differences can be tricky! So, basically, the attitude and devotion of NASA’s employees give us the last push to begin this journey in space-related research. We really needed that to expand our focus areas from healthcare and autonomous vehicles to space.

BBJ: Your solution to NASA proposed a deep learning model. How exactly does it work?

AOK & RAV: The core of the software is a deep learning model engineered and trained by us. It uses 28 data points as inputs and predicts all shock response spectrum (SRS) values separately. The data points are, among others, the strength and type of the shock, the distance between the shock and the measurement point, the type of affected materials, the type and number of joints, and the requested frequency. The model is surrounded by several functions that clean and normalize the input data. The final model is tiny, easy to run, and retrain for the parameters of other satellites or devices. The inference can be managed even in microcomputers.

BBJ: How can it be used in a real-life situation?

AOK & RAV: We think our software will have a tangible impact on the aerospace industry since the different components of satellites or spacecraft and their payloads may be damaged by various frequencies of the shock spectrum. A shock can be from a hit or an explosion. Both types of shock can happen during the launch and separation phases or during the space mission. High frequencies can harm the electronics, while lower frequencies can damage structural components. With the help of the solution, spacecraft components can be engineered more appropriately according to the expected shock response. More properly engineered devices lead to prolonged operation or less weight. Both results have a positive effect on the use of resources since, during a launch, every gram counts, and a longer functioning device needs replacing less often, which again means savings.

There are other possibilities in the long term. If we can reduce the mass of a spacecraft, it is possible to place lightweight shock sensors on critical points of the structure. Based on the live data feed, our model can easily be retrained to provide a flexible and adaptive shock prediction system. One day, a model like this may be a standard part of safety systems and mechanisms on a spacecraft.

BBJ: How will your solution help NASA specifically, and how might it affect the future of space flight?

AOK & RAV: Although the attributes of materials and joints are pretty assessable, tiny miscalculations can lead to significant errors. High computational capacity and deep learning can provide better models for shock propagation since, based on a few well-measured experiments, they can predict the probable shock propagation very well. Hopefully, in the not too distant future, it can be used in long-term deep-space flight.

We think NASA will use our solution for developing next-generation satellites and spacecraft. We also submitted a white paper and additional development information to help NASA experts develop our solution further if needed. We are open to NASA using our solution in new fields. We hope we have added something to the future exploration and conquest of space.

BBJ: What is next for you after winning the challenge?

AOK & RAV: Our deep learning model can be retained for new purposes like different structures or restructured based on the same concept, like handling a new type of joint mechanisms or new materials. Those processes are easy for us since we made the model and deeply understand it. We can save a lot of time and hard work for engineers if they ask for our help. So, we are waiting for a phone call or the notification beep for an incoming email from NASA. The concept of our idea can be used in different areas or industries. We are seeking challenges in our daily life, so we are looking both for new business possibilities on the market and new challenges on

The Aftershock NASA Challenge

The Aftershock challenge was run by the U.S. space agency,, the world’s largest freelancing and crowdsourcing marketplace by the number of users and jobs posted, and government-focused consultancy LMI. It aimed to crowdsource novel prediction models to improve the agency’s ability to predict shock loads through spacecraft.

Contestants from around the world were given four months to complete their entry providing a new model for shock propagation. The challenge received a total of 49 submissions. Of those, four solutions were awarded a share of USD 50,000.

This is the second NASA/ challenge Axel Ország-Krisz and Richárd Ádám Vécsey have won, having picked up USD 10,000 in the NASA Risky Space Business Challenge back in May of this year.

This article was first published in the Budapest Business Journal print issue of July 29, 2022.

Alteo Shareholders Approve HUF 400/Share Dividend Green Energy

Alteo Shareholders Approve HUF 400/Share Dividend

Bulgaria's Household Income, Spending Rise 20% in 2023 World

Bulgaria's Household Income, Spending Rise 20% in 2023

MBH Bank’s Post-merger HR Vision HR

MBH Bank’s Post-merger HR Vision

Silvanus Hotel Under Reconstruction, Will Reopen in Fall Hotels

Silvanus Hotel Under Reconstruction, Will Reopen in Fall


Producing journalism that is worthy of the name is a costly business. For 27 years, the publishers, editors and reporters of the Budapest Business Journal have striven to bring you business news that works, information that you can trust, that is factual, accurate and presented without fear or favor.
Newspaper organizations across the globe have struggled to find a business model that allows them to continue to excel, without compromising their ability to perform. Most recently, some have experimented with the idea of involving their most important stakeholders, their readers.
We would like to offer that same opportunity to our readers. We would like to invite you to help us deliver the quality business journalism you require. Hit our Support the BBJ button and you can choose the how much and how often you send us your contributions.