Gabriel Betton: Exploring the karst networks

What does the entrance to a karst network look like? Often, it resembles a peaceful water source, making it hard to imagine what lies beneath: a vast network of dozens of kilometers of tunnels, galleries, and cavities carved into limestone and buried deep underground. If scuba divers were to attempt to explore these bottomless wells, they would very quickly reach their physiological limits, and the narrowness of certain passages would prevent them from proceeding.

Yet, there is a crucial need to understand the intricacies of these karst networks known as aquifers (literally, “water-bearing”), and how they respond to precipitation and pollution.

As Gabriel Betton explains, “hydrogeologists seek to model and understand the behavior of these aquifers during extreme weather events, such as droughts or, conversely, torrential rains. This involves precise mapping of cave systems to better understand their structure and the behavior of the water flowing through them.”

These findings help predict the movement of water bodies as well as the consequences of potential pollutant discharges, and trigger the necessary alerts to prevent contamination. Another issue for karst systems located near the sea is the rise of seawater through a system of communicating vessels—again, in the event of drought or excessive exploitation of the resource.

“The robot must be able to map the network by determining its position—a problem that is already difficult in terrestrial or aerial robotics, but even more so when you’re underwater or underground, where neither GPS signals nor radio waves can pass through, all while operating with inevitably limited computational resources,” notes the doctoral student.

The robot’s roadmap will be simple: stay in the center of the passage it is exploring, remember the junctions it has encountered so as not to get lost on the way back, and go as far as it can while ensuring it has enough energy to return to its starting point and deliver its data.
 

Where many would have been tempted to throw in the towel in the face of so many conflicting requirements, Gabriel Betton decided to take inspiration from nature. He came up with the idea of a spindle-shaped robot, like a fish, and, taking the analogy to its logical conclusion, he wondered if it was possible to mimic their caudal fin.

“In the end, we arrived at a sort of torpedo propelled not by a propeller but by a bio-inspired fin. It’s much more maneuverable than a propeller, it consumes very little energy, and the entire control and actuation system fits inside a sealed cylinder, with a spring linkage leading to the fin.”