An automated mobile air quality (AQ) sensor array will provide high quality environmental data within the confined physical parameters of a space habitat. Project results, including generated data, will be used to develop algorithms for artificial intelligence (AI) which will ultimately automate monitoring of experiments as well as life support systems on the International Space Station (ISS), the Lunar Gateway, and beyond.

Initially, SPEC Sensors will demonstrate a platform for AQ monitoring in a form factor compatible with autonomous robots such as the Astrobee, currently in use aboard the ISS National Laboratory. In this phase, ground-based laboratory evaluations will be performed with a mobile prototype to address flight certification requirements necessary for integration with the current Astrobee fleet, and for safe delivery to the ISS National Lab by Magnitude.io. These experiments will also generate the training data for the proposed machine learning algorithms developed by the University of Missouri – St. Louis. The complexity of these algorithms will determine hardware requirements for the final phases of the project. For example, a reference array mounted on a mobile robot, in conjunction with remote fixed arrays will require a unique implementation of edge-computing methods and mesh-compatible hardware.

In Phase II, the platform will perform passive AQ monitoring from the payload bay of an Astrobee during sorties on the ISS. The temporal, spatial, and physical environmental data collected from these flights will generate real training data for machine learning development, and require minimal support from the station’s crew. Finally, in Phase III of this project, with updated software and hardware, this system will be provided to NASA and the ISS National Lab for integration into current and future operational systems. With the successful demonstration of this technology, we expect other needs will arise that can be solved with this AI enabled array.

Missions on the moon or Mars will benefit from this lightweight, low power, AI enabled AQ platform to ensure safety in vehicles, airlocks, and portable life support systems. This will provide tools necessary to measure the quality of air in a habitat, and to identify and locate potential hazards in a closed environment (e.g., identifying and locating electrical shorts generating ozone and soot).

Additionally, an educational low-cost version in classrooms can allow for student engagement through a NASA focused air quality and health curriculum.

This will enable applications such as aerial and ground-based robots designed to monitor remote and difficult to access environments (i.e. within coal mines, factories, or marine vehicles). The system presented here can be tailored to a specific industry, redesigned to predict and prevent accidents, by the careful selection of the sensors in the array, and the AI’s response to specific conditions.