CubeAID, The Team:
We are an interdisciplinary and international team that got together in order to join our efforts and ideas towards a common goal. We were able to communicate with ease, share our thoughts and all together formed an impressive synergy which allowed us not only to work efficiently but also enjoy our time working as a team and bond much more than we imagined.
- Barrionuevo, Nestor: I have a degree in Geographical Information Systems (GIS) from the National University of Tres de Febrero (UNTREF). I am a researcher in the area of hydrology at the National Institute of Agricultural Technology (INTA). My research is the analysis of satellite images and Geographical Information Systems and hydrological modeling. My interest in participating in NASA's SpaceApps is due to knowing new ideas about the use of satellite images and data available to solve problems of the local and global population.
- Barshap, Guy: I'm a first year PhD student of computer science in the subject of security and privacy of blockchain at Israel (Ben-Gurion university). I was interested in this challenge because I am fascinated by space and the opportunity to explore other planets. Furthermore I was inspired by an Israeli astronaut Ilan Ramon R.I.P, who died in the Columbia shuttle mission and I wanted to contribute with my technical skills to help those missions become safer. In addition, I deeply believe in collaboration with others and as I traveled to Argentina for my studies, I saw this Hackathon as a great opportunity to learn more about various exciting fields from the local tech community of Buenos-Aires.
- Burroni, Tomás: I'm a third year Astronautical Engineering student and work at my university designing electronics for scientific experiments aboard satellites. My passion for space, technology and challenges are what made me want to participate in this amazing opportunity aside from the fact that this is a great hands-on training in space systems engineering. I'm also interested in the opportunity of multidisciplinary teamwork that this event offers and the chance to let our creativity fly in order to find different solutions to serious problems.
- Cerino, Emmanuel: I’m a Bachelor in Graphic Design from Argentina. I work mainly as an illustrator and animator, creating explainer videos, infographics, journalistic illustrations and concept art for movies. I have a great passion for astrophysics, space exploration, science communication, art and science fiction. I believe good art must have good foundations. When I found out about the hackathon, I didn’t think it twice. I was looking forward to collaborate with people from other disciplines, as I think that creative solutions usually come from outside our comfort zones. My goal is to find the best visual and conceptual strategy to engage audiences in scientific matters.
- Mucanna, Camila: I'm an Astronautical Engineering student from Argentina. I wish to acquire a complete training in my study field, which I think is only possible when you are willing to take risks and commit on a project. The idea of combining my passion for astronautics, science development, informatics and design thrills me and motivates me to always be in the search for an idea to work on. I believe opportunities like the one NASA offers with Space Apps Challenge are unique as they allow people from diverse disciplines to encounter, learn from each other and complement in order to achieve a mission.
- Penida, Matías: I am a young mechatronics engineer. I work in smartcultiva as a electronics designer. I have always been interested in science and technology. I aspire to solve as many technological problems as I can in my lifetime. I am not afraid of challenges, but rather find motivation in them.
- Schvartz, Adrian: Astronautical Engineering student at the University of San Martín. My current inspiration is Elon Musk. I am an astronomy enthusiast, I am fascinated by planets and above all by the spacecraft in which we travel, Earth. My goal is to find peaceful solutions to the problems generated by the greedy wars of today. I am a lover of this planet and to continue living in it we need to find peace through technology and brotherhood. That is why I appreciate the hackathon, since it generates that union between people, I hope that this same thing will happen in the whole planet.
Spacecrafts (satellites in orbit, interplanetary ships and deep space probes) on service suffer from micro-meteoroid and orbital debris (MMOD) impacts quite frequently during their missions. Usually those impacts are small and do not mean a threat to the structure, but the sum of hundreds of impacts or just one mayor collision may be enough to ruin an entire mission, or even worse, threaten the lives of the crew.
These projectiles travel at speeds of between 10 to 14 km/s and measure around 50 µm to 2 mm, which means that they could easily carry more energy than a bullet. These characteristics make it impossible to track and deflect them, so an active prevention system could not be implemented.
In cases where there is no crew there is currently no solution to this problem; when there is one it is not always possible for them to go out and fix it by hand, so an automated solution must be found.
To design an autonomous free-flyer to inspect a spacecraft for damage from Micro-Meteoroid and Orbital Debris.
The CubeAID mission will be developed in one of the most defying environments known by humans: outer space. With no pressure, extreme temperatures, high levels of radiation and other variables that living organism from earth and common devices cannot withstand. For that reason, materials and technology applied in the mission must be of the highest levels available so that its features allow a long-term lifespan in an outer space environment.
Given the difficulty of developing an active prevention system, the CubeAID system is inspired on a search and rescue scheme. The multiple stages aspect of the project uses the concept of synergy often seen in nature, especially in ant colonies or bee hives. The biggest advantage that nature has over technology is the ability to learn, evolve and adapt, this is why we chose to use artificial intelligence for the second stage, the image processing. We imagined the first stage as a hawk, which observes from the distance to find a prey, using its eyes with a narrow field of view but great precision and focus. The third stage is inspired by arthropods, particularly on how spiders are able to grab onto almost any surface and walk without losing their grip; as well as this, from their ability to shoot webs we got the idea of the epoxy loaded needle. Finally, investigating on the way they hunt their prey we got the idea of attaching vibration sensors to the most critical parts of the spacecraft in order to estimate zones of possible impacts for further analysis with the other stages.
Architecture and design:
A two-units cubesat orbiting the host vehicle collects information on the state of their external surface by taking pictures. This can be achieved due to a slight phase shift between the main vehicle’s orbit and the cubesat’s orbit around earth. Having orbits with the same orbital period (this is done by having the same semiparameter), same (or similar) eccentricity, and slightly different argument of perigee, true anomaly and inclination it is possible to achieve an orbit of the cubesat relative to the host vehicle that covers all its surface using minimal fuel for orbit mantainance. This orbit design allows for the cubesat to scan the entire surface in around 5 to 8 orbits, which in LEO takes less than a day. The orbit was specially designed by us using our own orbit simulator in python.
- 3 Flywheels for the Attitude Control Subsystem (ACS).
- 1 Pulsed Plasma Thruster for the Orbit Control Subsystem (OCS). The use of an electric propulsion thruster allows a more precise orbit control and a longer lifetime with significantly less fuel.
- Infrared and multi-spectral camera as the Main Payload (PL).
- 4 Solar panels, battery and power distribution as the Electrical Power Subsystem (EPS).
- UHF antenna to communicate with the host vehicle (Telecommunications Subsystem - TCS).
- Aluminum 7075 frame as the Structural Subsystem (SS).
- On board computer to control the rest of the subsystems, ACS to aim the camera, PL to take the images, TCS to send them to the main vehicle, and OCS in combination with ACS for orbital station-keeping.
We borrowed code built for bullet impact detection algorithms given the similarities between the two. The algorithm uses Color Segmentation, which is a technique that convert a RGB Image to a binary image (black and white) by dynamically detecting the major color in the image. Then the noise of the image is reduced by doing morphological operation (a term that is used in Image processing), that is to spread color smoothly. After this preprocessing is done, with the use of IBM’s Watson Visual Recognition we are able to distinguish the impact holes from the rest of the structure.
An Improvement to this algorithm, will be done by using a Machine learning detector algorithm. By producing and taking pictures of impacts and also using NASA’s Space Debris Sensor, which is functioning in the ISS since December 2017. The method will be based on cutting-edge algorithm for detection such as: FASTER RCNN or MASK RCNN.
Once the impacts are found, the specific coordinates are calculated (from a model of the vehicle) and given to the Spider.
A robot designed to imitate a spider with articulated extremities and exchangeable ends to make the spider more suitable for the surface it will be inspecting.
- 6 Fully articulated extremities.
- 6 shoe-like electromagnet pads.
- 6 shoe-like artificial setule-based nanostructure pads.
- Digital Microscope
- Camera with IR and RGB capabilities.
- Epoxy loaded needle.
The technologies used in its limbs, particularly in the pads, allow the spider to walk anywhere in the outer surface of the vehicle for an in-depth specific analysis of each impact zone. As well as this, it contains a needle capable of depositing an epoxy resin over the damaged areas in order to prevent the microfractures from spreading and causing major structural damage.
The two proposed technologies for the pads are: electromagnets and setule based nanostructures. Electromagnets would be more convenient on metallic surfaces. On the other hand, for ceramics or other surfaces, a nanostructure resembling the setules on the feet of spiders (these work on Van der Waal forces) would be the used.
Concept of Operations:
Stage 1 - Cubesat:
The cubesat performs an active surveillance and inspection of the exteriors of the spacecraft. It takes high definition images which are instantly sent to the main vehicle, where analysis is held. If the amount of images and data exceeds the capability of the vehicle, the data can be sent to a ground station where further analysis can be done.
Stage 2 - Image Processing
Onboard the main vehicle (with a higher computing power) or on ground, the images are processed in order to find places of interest where impacts are likely to be found. Once these places are pinpointed, the coordinates are sent to the spider.
Stage 3 - Spider Robot
Once the Spider Robot gets new coordinates it is released to approach the designated area, and with the help of a front camera it positions itself on the desired spot.
A digital microscope is activated and starts collecting precise data (depth, width, propagation) on the impact and characterizing the damage.
If considered suitable, an epoxy dose is disposed on the place to avoid future fracture propagations.
Once it finished it returns to its charging pod on the surface of the ship.
- We would like to develop and array of sensors, that could be deployed inside the spaceship in critical zones such as windows, hatches or special equipment. These sensors would measure the vibrations and deformations when an impact occurs near those areas. Then the cubesat would focus on taking more pictures of that specific zone in order to add another layer of security against critical failures.
- Develop a multi-modal system for the spider’s legs that allows to switch between pads with different technologies so that the robot best adapt to the explored surface.
- Coordinated swarm flight. In interplanetary missions or deep space probes it is not possible to orbit around the main vehicle due to the extremely low attraction between bodies (and the designed orbit does not work without the earth, or another celestial body exerting a considerable gravitational force on them). Therefore, with an arrangement of four cubesats in a tetrahedron shape, with the main vehicle in the center, it would allow a full surface coverage with minimal fuel use. When the main vehicle is not using the thrusters it can be used as an inertial center of coordinates, so the cubesats would be “still” in specific locations.
- Development of an application that allows for people in Earth to see the pictures taken by the cubesat and help in the search of impact zones. This would help the artificial intelligence used for the image processing learn faster.
Having developed the Pre-Phase A of our mission here is a summary of the key points of the development:
- We have designed an integral solution combining three separate systems, that work together as a team and are therefore able to accomplish much greater results. The cubesat provides a fast periodic analysis of the whole ship, which would be impossible to do using only the spiderbot; and the spiderbot provides an in depth analysis of each spot and a patch to prevent further damage, which again would be impossible for the cubesat to do.
- There is no need to develop new technology since everything that we used already exists. All the subsystems and parts in the cubesat are available in the market; the machine learning algorithm is based on algorithms used for recognizing bullet holes; the two proposed technologies for the spiderbot's pads already exist (the setules one is still being researched but successful prototypes have been created).
- We were able to find the right orbital parameters for the cubesat so that it appears to orbit around the ISS (the same method can be easily applied for any other satellites only by changing some numbers in the code). The orbit was also simulated in our own python code in order to validate our design.
- We have made a great effort in using our systems engineering knowledge to research all of the aspects of the mission (not only the subsystems in each stage but also the interfaces) to make sure our system has a solid and robust design and validate all of our proposals. Therefore we present something that could actually be built and put into work in a relatively short time and with limited resources.
- The system as a whole is scalable since more cubesats or spiderbots could be added to work in parallel and have higher response speeds. As well as this, each part is easily replaceable should there be a critical problem with any stage.
- Finally, we have left the possibility open to keep developing the project and adding more features since we recognize that it can still be improved and worked on.
- “Fundamentals of Astrodynamics and Applications”, David A. Vallado. 3rd edition, 2007, Space Technology Library. For theory on orbits and algorithms for orbit simulations.
- “Preliminary Development of a Bio-Inspired Hexapod Climbing Robot relying on Dry Adhesives”, Yasong Li, Master of Applied Science Thesis. Simon Fraser University, 2009. Technology proposed for use in the spider’s pads.
- NASA Systems Engineering Handbook, 2007.
- "Fuel consumption and collision avoidance strategy in multi-static orbit formations", E.F. Jochim, H. Fiedler, G. Krieger. Microwaves and Radar Institute, German Aerospace Center (DLR), Oberpfaffenhofen, Germany. Oct 16, 2010.
- In-Space Inspection Technologies Vision Workshop, 2012.
- https://www.space.com/38984-tiny-space-debris-sensor-to-station.html NASA’s Space Debris Sensor
- Stackoverflow for most of the python problems or doubts.