![](/style/images/tocco.png)
Gabriele Gennaro
Autonomous recognition and pose estimation in dusty environment for space application.
Rel. Marcello Chiaberge. Politecnico di Torino, Corso di laurea magistrale in Mechatronic Engineering (Ingegneria Meccatronica), 2020
|
PDF (Tesi_di_laurea)
- Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (22MB) | Preview |
Abstract: |
With regards to space explorations, significant achievements have been realized during the last 30 years, thanks to advancement in robotic sector . In particular, Artificial Intelligence development allowed several challenging missions and supported scientist and astronauts in the spacecraft control, autonomous navigation or in object detection tasks. In 2018, NASA provided 330k $ to a researchers team, aiming to develop an intelligent system able to navigate amid the debris, detecting and avoiding them . During same year a partnership between NASA and Intel helped astronauts during their training phase for their mission. Researchers trained a neural network with thousands of Moon images in order to create a virtual moon environment and compare it with the local environment . Nowadays one of the most significant space challenge for humanity is Mars exploration. In 1965 Mariner 4 performed a movement called flyby, passing in the vicinity of Red Planet and capturing a close-up picture of it. That was the very first picture of Mars from so close. The Mariner was a robotic craft of small size and its purpose was to explore Mars,Venus and Mercury. During the years, martian mission have become more and more sophisticated , driven by incredible technological advancement especially in autonomous system. Successful examples of how intelligent machines play a pivotal role in space exploration are the Spirit and Opportunity rovers. Involving object detection algorithm and autonomous navigation, mars rovers have managed to perform geological analyses trough their portable laboratories on the spot . It is considered one of the most successful mission for NASA . The evolution of the study of Mars characteristic has lead to “Mars Sample Return” mission. The scope of this mission is to pick samples , pre-filled with martian soil and gas, and bring them to Earth, in order to analyse samples with more complex instruments than the previous mission. This project represents a fundamental step towards an overall comprehension of the past history of Mars and the possibility to contemplate a human colonization. In this thesis project, an autonomous recognition system is proposed as possible solution for the picking task of the MSR rover. The system has to detect the sample and perform a pose estimation in order to collect information for the picking phase. Technological limits related to spaceflight and Mars environment factors have been taken into account aiming to create a system designed for a real application . |
---|---|
Relatori: | Marcello Chiaberge |
Anno accademico: | 2019/20 |
Tipo di pubblicazione: | Elettronica |
Numero di pagine: | 86 |
Soggetti: | |
Corso di laurea: | Corso di laurea magistrale in Mechatronic Engineering (Ingegneria Meccatronica) |
Classe di laurea: | Nuovo ordinamento > Laurea magistrale > LM-25 - INGEGNERIA DELL'AUTOMAZIONE |
Ente in cotutela: | ETSI INDUSTRIALES - UNIVERSIDAD POLITECNICA DE MADRID (SPAGNA) |
Aziende collaboratrici: | Thales Aliena Space |
URI: | http://webthesis.biblio.polito.it/id/eprint/15269 |
![]() |
Modifica (riservato agli operatori) |