SNAKE - Control Strategies for Snake Robot Locomotion in Challenging Outdoor Environments

The use of future service and security robots in e.g. search and rescue, pipe inspection, and firefighting operations will rely on their ability to maintain mobility in unknown and challenging environments. Robots are also important in exploring planets and the Moon, as well as in connection with operations of infrastructure in space. Snake robots carry the potential of contributing significantly in such applications due to their long, slender and flexible body that can provide robust propulsion skills in virtually any environment. However, to realize this potential, snake robots need to be able to tackle a wide range of different environments. The SNAKE project have targeted research challenges imposed by snake robot locomotion in cluttered and "application-realistic" environments. Inspired by biological snakes, external objects and irregularities are considered beneficial since they represent push-points that the snake robot can curl around in order to push its body forward, i.e. perform 'obstacle-aided locomotion'. This inherently robust form of locomotion has been a strategic research focus at NTNU and SINTEF since 2004. The overall objective of SNAKE was to realize obstacle-aided locomotion in cluttered outdoor environments. To this end, the SNAKE project has produced stepping stones for realizing such locomotion with, e.g., 1) a thorough introductory literature review as a starting point for researchers and industry where we focus on perception-driven obstacle-aided locomotion - i.e., where a snake robot can use sensory input to better perform obstacle-aided locomotion, 2) SnakeSIM - a simulator framework with simulated sensors included and integrated with the Robot Operating System (ROS) where new algorithms can be more easily tested to facilitate rapid development, 3) a modeling and control approach with a concept called Virtual Functional Segmentation to greatly simplify designing new control algorithms for snake robots in cluttered outdoor environments. The approach makes it possible to focus more on the overall goal of snake robot locomotion, rather than having to deal with the complexity of controlling each snake robot joint individually, and 4) local path planning for snake robots where we provide criteria and an optimization system for snake robots to determine which obstacle (or "push-point") is best to use next to move toward a desired goal location. These steppingstones may impact both the research community as well as the industry. The SNAKE project came to an end in mid-2019, but the researchers involved and the research groups seek to continue on the journey toward fully functional snake robots that can benefit the industry and the society at large.

I SNAKE-prosjektet har vi tatt forskningsfeltet innen hindringsbasert fremdrift med slangeroboter videre og utviklet nye byggesteiner innen feltet. En mulig virkning av dette er at resultatene kan brukes - både innen forskning og innovasjon - for å utvikle hindringsbasert slangerobotfremdrift videre og komme nærmere kommersielt utnyttbare slangeroboter. Hvis kommersielle aktører velger å ta utgangspunkt i resultatene fra prosjektet så kan dette føre til nærmere samarbeid mellom FOU-miljøene og industrien. Synligheten med publisering på den internasjonale forskningsarenaen kan føre til økt internasjonalt forskningssamarbeid. En mulig langsiktig effekt er at resultater fra SNAKE blir en del av løsningen for å kunne realisere de mulighetene som ligger i å ta i bruk slangeroboter innen områder der slangeroboter har potensiale for å bidra - f.eks. innen beredskap, inspeksjon og vedlikehold og innen romfart (bl.a. utforske planeter).

The use of future service and security robots in e.g. search and rescue, pipe inspection, and firefighting operations will rely on their ability to maintain mobility in unknown and challenging environments. Snake robots carry the potential of contributing significantly in such applications due to their long, slender and flexible body that can provide robust propulsion skills in virtually any environment. However, practical applications of these mechanisms are still very limited since the research communities have so far focused primarily on control strategies for locomotion over flat surfaces. The SNAKE project targets research challenges imposed by snake robot locomotion in cluttered and "application-realistic" environments. Inspired by biological snakes, external objects and irregularities are considered beneficial since they represent push-points that the snake robot can curl around in order to push its body forward, i.e. perform 'obstacle-aided locomotion'. This inherently robust form of locomotion has been a strategic research focus at NTNU/SINTEF over the last years. The overall objective of SNAKE is to realize obstacle-aided locomotion in cluttered outdoor environments. The project will develop new control strategies for obstacle-aided locomotion with focus on environment adaptation and energy efficiency. The project will experimentally investigate the control approaches using a new and unique snake robot prototype recently developed by SNAKE researchers. The project will pursue collaborations with AMOS (a Norwegian Centre of Excellence) and the two upcoming EC projects TRACT and INACHUS, where SINTEF is a central partner. Project results will be disseminated both nationally and internationally.