Let's Space It. Mikael Granit

In this series of articles, we are introducing our colleagues behind ReOrbit technology. What brought them to the industry, their journeys towards humanity's ongoing expansion across the final frontier, lessons learned, and things they’d create if there were no technological limitations.

Introducing Mikael Granit, a Machine Learning Engineer at ReOrbit, who envisions artificial intelligence as a key ally in addressing the escalating challenge of space debris and prolonging the lifespan of satellites. Mikael's fascination with science and enchantment of space trace back to his childhood—he vividly recalls engaging in lengthy phone discussions about planets with his grandmother even at the tender age of 5, highlighting his innate draw towards the marvels of the cosmos. Read on to find out how Mikael turned his childhood dream into a full-time job.

‍Why did you choose space

I began my journey in this captivating field, with the study of astronomy, and later transitioned into a career within the industry. My passion for space dates back to childhood, and what continues to captivate me is the challenging and 'forbidden' nature of the environment—a realm not naturally meant for human exploration. Nevertheless, it's a testament to the stubbornness of human curiosity that we insist on exploring these extremely hostile environments. To be part of this endeavour has been a dream of mine for a very long time.

You started your career as

Choosing to study physics at Åbo Akademi University was driven by my keen interest in astronomy. I appreciated that the program offered a spectrum of other physics disciplines, enhancing my academic experience. During my studies, I discovered programming and machine learning, which I enjoyed a lot. I focused on that as it seemed like a very promising career path. I got the opportunity to apply my skills for a Master's project with the Material Science Department at Turku University, where I learned software development in a professional setting. When it came time to look for a job, I very much based it on my deep-seated passion for space and, thus, found my current position at ReOrbit on LinkedIn. There was an opening for a machine learning engineer, and I thought this would be a dream come true to contribute my educational experience in machine learning and programming to the space sector. Luckily, I got the job.

Your most memorable memory so far

It's not necessarily technical but more related to the workplace. The atmosphere at ReOrbit is dynamic and enriched with people from all over the world sharing different experiences and cultures. One time I had a chance to introduce the Finnish sauna culture to my international colleagues. Another fond memory is a karaoke night which eventually led to a quartet of Argentinians singing ‘’Despacito’’.

Your most valuable lesson so far

You will always learn the most by talking to people knowledgeable in a subject rather than just reading research papers and books. Communicating with knowledgeable people is better than solely relying on research.

The one thing that makes you smile

The people here are very good at making others smile. I think the passionate atmosphere created by people working together towards something exciting, like a launch, contributes to a positive and engaged work environment. Everyone here is working towards something they find very exciting and that brings me a joyful smile.

What does nobody prepare you for‍

The amount of research required for this line of the industry. I came in with a slightly naive sense of a problem being solved due to the thousands of satellites already floating in space, and not considering the innovation potential. A small startup like ours requires staying at the cutting edge of innovation to pioneer the transformation.‍

Imagine there are no technological limitations, what do you wish you could design/create in the space sector

Well, there are many different areas where machine learning could be applied right now. Specifically, in the orbital determination and collision avoidance part of satellites. Satellites are very expensive and challenging to repair, so the problem of junk in space and large-scale constellations calls for a solution.

AI is suitable for making satellites more autonomous and aware of their surroundings, which, therefore, allows them to communicate with each other to perform orbital manoeuvres and reduce the risk of collisions. That would be the ideal scenario, especially for large satellite constellations. Furthermore, there should be standardised protocols to avoid collisions, like how roads have rules to prevent accidents. This is a problem that needs to be solved, and I believe ReOrbit's innovation can contribute to a solution.