Sometimes when terrain is inaccessible to humans we use robots to help us explore it. In this project a self navigating robot was created that used ultrasonic sensors to detect obstacles in its path and avoided them. When an obstacle is encountered the robot documents said obstacle by taking photographs of it as well as registered its coordinates and created a map, as a bonus feature. Thereafter the robot continued its path based on where obstacles were absent. By using stepper motors to drive the robots traveled distance was calculated. With this information a map over the traveled path was created. Tests were conducted where the map was compared to real life as well as letting the robot roam freely. The tests showed the robots ability to evade obstacles and how well the integrated camera function performed. The placement of the sensors worked well enough considering only five were used. Although the robot would improve significantly if an increased amount of sensors were to be added. The algorithm enabled the robot to navigate and avoid all detected obstacles. It is the sensors that inhibited its navigation since they only detected obstacles directly in front of them. Since this was a mobile robot it was powered by batteries. The robot would be able to explore to a greater extent if it could recharge its batteries on its own, for example with solar panels. A GPS could be installed to keep track of the robot at all times. / Ibland är okänd terräng av olika skäl otillgänglig för människor och då kan en robot istället skickas ut för att undersöka dessa omgivningar. I detta projekt har en självnavigerande robot skapats där ultraljudssensorer användes för att registrera hinder som sedan dokumenterades. Dokumentationen gjordes genom att fotografera hinder och som en extra funktion registrerades deras koordinater och med dessa skapades en karta. Därefter valdes en ny riktning baserat på vilken väg som var fri från hinder. Stegmotorer användes för att driva roboten och genom att räkna på de antal steg dessa tog beräknades det avstånd roboten hade förflyttat sig. Med hjälp utav dessa avstånd skapades en karta över den tillryggalagda vägen. Tester utfördes där kartan jämfördes med verkligheten samt låta roboten upptäcka på fri hand. Detta visade hur väl den undvek hinder och hur väl den integrerade kamerafunktionen fungerade. Sensorernas placering fungerade bra med tanke på att endast fem användes. Däremot kan fler adderas för en mer exakt avläsning. Algoritmen fungerade väl då roboten parerade alla upptäckta hinder. Det var endast sensorerna som hämmade robotens navigering då de bara upptäckte hinder direkt framför dem. Då roboten var mobil behövdes batterier för att ge den ström. En framtida förbättring innefattar därför exempelvis att addera en solpanel som kan ladda batterierna för att göra den ytterligare självständig. Dessutom kan en GPS installeras för att kunna bevaka var roboten befinner sig i alla situationer.

This thesis describes the design, implementation and analysis for a preliminary study for DLR's MASCOT lander's next mission to Small Solar System Bodies (SSSB). MASCOT (Mobile Asteroid Surface Scout) is a nano-lander that flew aboard Hayabusa2 (JAXA) to an asteroid, Ryugu. It is a passive nano-spacecraft that can only be deployed ballistically from a hovering spacecraft. Current research focusses on optimizing similar close-approach missions for deploying landers or small cubesats into periodic orbits but does not provide solutions with semi-autonomous small landers deployed from farther distances. This study aims to overcome this short-coming by proposing novel yet simple Guidance, Navigation and Control (GNC) and Propulsion systems for MASCOT. Due to its independent functioning and customisable anatomy, MASCOT can be adapted for several mission scenarios. In this thesis, a particular case-study is modelled for the HERA (ESA) mission. The first phase of the study involves the design of a landing trajectory to the moon of the Didymos binary asteroid system. For a preliminary analysis, the system - Didymain (primary body), Didymoon (secondary body) and MASCOT (third body) - are modelled as a Planar Circular Restricted Three Body Problem (PCR3BP). The numerical integration methodology used for the trajectory is the variable-step Dormand–Prince (Runge Kutta) ODE-4,5 (Ordinary Differential Equation) solver. The model is built in MATLAB-Simulink (2019a) and refined iteratively by conducting a Monte Carlo analysis using the Sensitivity Analysis Tool. Two models - a thruster-controlled system and an alternative hybrid propulsion system of solar sails and thrusters - are simulated and proven to be feasible. The results show that the stable manifold near Lagrange 2 points proposed by Tardivel et. al. for ballistic landings can still be exploited for distant deployments if a single impulse retro-burn is done at an altitude of 65 m to 210 m above ground with error margins of 50 m in position, 5 cm/s in velocity and 0.1 rad in attitude. The next phase is the conceptual design of a MASCOT-variant with GNC abilities. Based on the constraints and requirements of the flown spacecraft, novel GNC and Propulsion systems are chosen. To identify the overriding factors in using commercial-off-the-shelf (COTS) for MASCOT, a market survey is conducted and the manufacturers of short-listed products are consulted. The final phase of the study is to analyse the proposed equipment in terms of parameter scope and capability-oriented trade-offs. Two traceability matrices, one for devised solutions and system and another for solutions versus capabilities, are constructed. The final proposed system is coherent with the given mass, volume and power constraints. A distant deployment of MASCOT-like landers for in-situ observation is suggested as an advantageous and risk-reducing addition to large spacecraft missions to unknown micro-gravity target bodies. Lastly, the implications of this study and the unique advantages of an enhanced MASCOT lander are explored for currently planned SSSB missions ranging from multiple rendezvous, fly-by or sample-return missions. Concluding, this study lays the foundation for future work on advanced GNC concepts for unconventional spacecraft topology for the highly integrated small landers. / <p>This thesis is submitted as per the requirements for the Spacemaster (Round 13) dual master's degree under the Erasmus Mundus Joint Master's Degree Programme. </p> / MASCOT team, DLR