About

Our team met online with Brian Ringley, the Construction Technology Manager at Boston Dynamics, to get his expert insight on the application of robotics sensing in industrial settings. Ringley is one of the main developers of the dog-shaped mobile sensor platform SPOT. Not only is SPOT a technological marvel but it can be deployed in sensing and monitoring industrial environments, such as construction sites or chemical plants. SPOT uses thermal imaging and vibration monitoring to look for anomalies and alert humans. SPOT's inspection can alleviate the exposure risks of workers from treacherous terrain, explosives, and hazardous gas.

EXTRACTS

Fig 1: Brian Ringley

Implementation

We were interested in how Ringley would implement our bacteria-electronics interface in the SPOT platform if he could. He proposed that we think of the ways animals sense to design our robots. Combining biology with robotics could imitate the sense of "smell" or "taste". He saw the monitoring potential for a bacterial biosensor component in SPOT that could investigate hazardous gas, toxic industrial leakage, or offshore oil rigs malfunction, for example. The advantage of hosting a monitoring process in a mobile robot is that you can send these robots and collect data in environments that are hostile or expensive for humans to go to. He sees a lot of potential for mobile robots to explore and collect data in toxic areas, warzones, aquatic environments, space, or areas with extreme weather conditions. To quote Ringley, "robots are most valuable when they can do something in combination with humans, something that robots and humans can't do on their own".

Beyond the sensing scope, mobile robots collect data in a repetitive manner at the same place, every day, at the same time, and therefore construct a record of data to detect trends. New means of data collection combined with the automation of analysis tools could greatly impact our understanding of the world around us. We can envision a future where our increased perception and understanding of the world can help us manage the challenges and dangers that we and the planet face.

Fig 2: SPOT in actions

Safety of autonomous robots

When asked about the dangers of robots gaining autonomy from people, Ringley argued strongly against those concerns. He assured that there is always someone controlling the robot. SPOT, for example, cannot venture outside of its assigned map, and that eventually its battery would run out without human intervention. He says that robots becoming sentient is a fear stemming from science fiction. He proposes that we shed the fear of robots replacing humans and instead try to create robots to accompany humans: "robots should allow us to do things". He assures that robots are not the most dangerous and that other automation technology, such as algorithms on social media, should be a stronger focus for concern.

Integration

Since we met Ringley early on in our project, we were able to integrate his insights into the design of our project. We decided our hardware had to be modular and programmable so we could automate our experiments as much as possible and increase repeatability. We integrated a biosensor system, based on peaks of conductivity measures induced by a chemical, to demonstrate an investigation and monitoring potential as an application of our toolkit. In addition, the activation of gene expression by an electrical signal in our system reproduces the feedback system observed in robotics, which would allow a two-way communication between the bacteria and the host robot.

In nature, we observe symbiosis on many levels of complexity. Single-celled photosynthetic algae live in symbiosis with corals which provide a macro structure conferring protection. Ringley suggested that we get inspired by animals, so we conceptualized a symbiotic relationship between single-celled bacteria hosted in the macro architecture of a robot. The bacteria provide the robot with data on the chemical world around, and the robot allows actuation from this data in motility and signal processing.