Engineering a robotic approach to mapping exposed volcanic fissures

Abstract:

Field geology provides a framework for advanced computer models and theoretical calculations of volcanic systems. Some field terrains, though, are poorly preserved or accessible, making documentation, quantification, and investigation impossible. Over 200 volcanologists at the 2012 Kona Chapman Conference on volcanology agreed that and important step forward in the field over the next 100 years should address the realistic size and shape of volcanic conduits. The 1969 Mauna Ulu eruption of Kīlauea provides a unique opportunity to document volcanic fissure conduits, thus, we have an ideal location to begin addressing this topic and provide data on these geometries. Exposed fissures can be mapped with robotics using machine vision. In order to test the hypothesis that fissures have irregularities with depth that will influence their fluid dynamical behavior, we must first map the fissure vents and shallow conduit to deci- or centimeter scale. We have designed, constructed, and field-tested the first version of a robotic device that will image an exposed volcanic fissure in three dimensions. The design phase included three steps: 1) create the payload harness and protective shell to prevent damage to the electronics and robot, 2) construct a circuit board to have the electronics communicate with a surface-based computer, and 3) prototype wheel shapes that can handle a variety of volcanic rock textures. The robot’s mechanical parts were built using 3d printing, milling, casting and laser cutting techniques, and the electronics were assembled from off the shelf components. The testing phase took place at Mauna Ulu, Kīlauea, Hawai’i, from May 5 – 9, 2014. Many valuable design lessons were learned during the week, and the first ever 3D map from inside a volcanic fissure were successfully collected. Three vents had between 25% and 95% of their internal surfaces imaged. A fourth location, a non-eruptive crack (possibly a fault line) had two transects imaging the textures of the walls with depth to compare to the fissure vents. The vent surface irregularity documented by Parcheta et al., (accepted) continues with depth into the fissure; depths are variable depending on the amount of talus filling them. We will show and discuss data from the main field vent location, and movies of descents in to the other vents.