An intelligent algorithm for autonomous scientific sampling with the VALKYRIE cryobot

Abstract

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.

Microbial processes in the weathering crust aquifer of a temperate glacier

“…Meltwater samples from discrete ice depth intervals were obtained using a customized water sampler and filtration system that was integrated as a scientific payload on the cryobot (Clark et al, 2017;Stone et al, 2018). Operation of the cryobot requires a water-filled cavity; therefore, the 30 cm aluminum melt head was used to create a ∼ 1 m pilot hole, the meltwater generated was replaced with an equal volume from a supraglacial stream, and this water was continually circulated through a 0.2 µm filter for ∼ 1 h to reduce the number of microorganisms in the priming water.…”

“…Annual melt cycles generate water on the surfaces of glaciers and ice sheets that is transported through a sinuous network of channels to the margin as supraglacial runoff or glacier bed via crevasses and moulins to be discharged subglacially (Smith et al, 2015;Chu, 2014). Meltwater can also be temporarily stored on the surface in cryoconite holes (Hodson et al, 2008;Edwards and Cameron, 2017), supraglacial lakes (Fitzpatrick et al, 2014;Langley et al, 2016), and nearsurface aquifers within the ice (Hoffman et al, 2014;Karlstrom et al, 2014;Cook et al, 2016). The low-density, highly porous surface ice prevalent in ablation zones serves as a reservoir for meltwater, affecting the timing and magnitude of its delivery to supraglacial, subglacial, and proglacial systems (Munro, 2011;Smith et al, 2017;Cooper et al, 2018;Stevens et al, 2018).…”

“…In the ablation zone, the absorption of shortwave radiation within the ice promotes internal melting and density reduction (LaChapelle, 1959;Munro, 1990), forming a highly porous veneer on the glacier surface termed the weathering crust (Müller and Keeler, 1969). Flow paths through the weathering crust's saturated water table link segments of the supraglacial hydrological system, transporting microbes and nutrients via subsurface flow (Irvine-Fynn et al, 2012;Hoffman et al, 2014;Karlstrom et al, 2014;Cook et al, 2016). The availability of photosynthetically active radiation (PAR) in the near-surface aquifer may support phototrophic metabolism (Hodson et al, 2013;Irvine-Fynn and Edwards, 2014), allowing the development of microbial communities in the glacial weathering crust during the melt season.…”

“…The availability of photosynthetically active radiation (PAR) in the near-surface aquifer may support phototrophic metabolism (Hodson et al, 2013;Irvine-Fynn and Edwards, 2014), allowing the development of microbial communities in the glacial weathering crust during the melt season. The weathering crust aquifer (WCA) is presumed to be an important component of the hydrological and biogeochemical budget for cold-based glaciers (Irvine-Fynn et al, 2012;Cook et al, 2016;Rassner et al, 2016), but no data have been available to assess microbial activities and biogeochemical transformations in temperate ice systems.…”

Exo-Ocean Exploration with Deep-Sea Sensor and Platform Technologies

“…Active exploration could be performed while penetrating the ice, in order to detect remnants of life that were frozen in the ice when the shell ruptured. In particular, the VALKYRIE cryobot had an early on-board meltwater sampling system and an autonomous algorithm to command sampling (Clark et al, 2017).…”