Using visual recognition and machine learning for the sorting of potential 3D-printable materials in space. NASA OTW, Best Pr., Rookie, Isth
As we move forward with space exploration, it is important to reduce our carbon footprint. 3D printing in space is a more efficient way of developing tools without having to constantly send materials from earth to space. Now, astronauts can make their own tools using NASA’s newly developed 3D printer “Refabricator” that uses recycled materials such as metal and plastic.
We believe that automation is the key to an efficient future and thus being able to cut the time of a process short even by a small amount of time is a valued effort. Our Space Auto Sorter for Sustainability or “SASS” for short will use vision recognition and machine learning software based on a Raspberry Pi to sort through materials that can be recycled for 3D printing purposes and make the process more efficient.
In today’s industry, machine learning has become more popular at fabrication lines. The technology is not perfect, however, with Elon Musk’s Tesla car company having a difficult time meeting production demand and Elon attributing this holdback to “too much automation”. The training and generalization of machine learning models still continue to be one of the main issues of this technology. Being able to find the best training algorithm that effectively separates data in the right groups has not been perfected yet.
We hope to one day see a world that uses artificial intelligence for the benefit of humanity regardless of social status, race or sexual orientation. It is with this mindset that we continue to move forward as we see researchers continue to work on the development of this technology.
Model Sorter CAD
Current Project / Post can also be found using: