The ARTEMIS docking system demonstrates autonomous docking capability applicable to robotic exploration of sub‐ice oceans and sub‐glacial lakes on planetary bodies, as well as here on Earth. In these applications, melted or drilled vertical access shafts restrict vehicle geometry as well as the in‐water infrastructure that may be deployed. The ability of the vehicle to return reliably and precisely to the access point is critical for data return, battery charging, and/or vehicle recovery. This paper presents the mechanical, sensor, and software components that make up the ARTEMIS docking system, as well as results from field deployment of the system to McMurdo Sound, Antarctica in the austral spring of 2015. The mechanical design of the system allows the vehicle to approach the dock from any direction and to pitch up after docking for recovery through a vertical access shaft. It uses only a small volume of in‐water equipment and may be deployed through a narrow vertical access shaft. The software of the system reduces position estimation error with a hierarchical combination of dead reckoning, acoustic aiding, and machine vision. The system provides critical operational robustness, enabling the vehicle to return autonomously and precisely to the access shaft and latch to the dock with no operator input.

No abstract

ABSTRACT. VALKYRIE (Very-deep Autonomous Laser-powered Kilowatt-class Yo-yoing Robotic Ice Explorer) is a NASA-funded project to develop key technologies for an autonomous ice penetrator, or cryobot, capable of delivering science payloads through outer planet ice caps and terrestrial glaciers. This 4 year effort will produce a cylindrical cryobot prototype 280 cm in length and 25 cm in diameter. One novel element of VALKYRIE's design is the use of a high-energy laser as the primary power source. 1070 nm laser light is transmitted at 5 kW from a surface-based laser and injected into a customdesigned optical waveguide that is spooled out from the descending cryobot. Light exits the downstream end of the fiber, travels through diverging optics, and strikes an anodized aluminum beam dump, which channels thermal power to hot-water jets that melt the descent hole. Some beam energy is converted to electricity via photovoltaic cells, for running on-board electronics and jet pumps. Since the vehicle can be sterilized prior to deployment, and forward contamination is minimized as the melt path refreezes behind the cryobot, expansions on VALKYRIE concepts may enable cleaner access to deep subglacial lakes. This paper focuses on laser delivery and beam dump thermal design.

The development of algorithms for agile science and autonomous exploration has been pursued in contexts ranging from spacecraft to planetary rovers to unmanned aerial vehicles to autonomous underwater vehicles. In situations where time, mission resources and communications are limited and the future state of the operating environment is unknown, the capability of a vehicle to dynamically respond to changing circumstances without human guidance can substantially improve science return. Such capabilities are difficult to achieve in practice, however, because they require intelligent reasoning to utilize limited resources in an inherently uncertain environment. Here we discuss the development, characterization and field performance of two algorithms for autonomously collecting water samples on VALKYRIE (Very deep Autonomous Laser-powered Kilowatt-class Yo-yoing Robotic Ice Explorer), a glacier-penetrating cryobot deployed to the Matanuska Glacier, Alaska (Mission Control location: 61°42′09.3″N 147°37′23.2″W). We show performance on par with human performance across a wide range of mission morphologies using simulated mission data, and demonstrate the effectiveness of the algorithms at autonomously collecting samples with high relative cell concentration during field operation. The development of such algorithms will help enable autonomous science operations in environments where constant real-time human supervision is impractical, such as penetration of ice sheets on Earth and high-priority planetary science targets like Europa.

No abstract