Some time after 2024, routine travels to the moon could be an ‘affordable’ reality, thanks to AI and machine learning technology.
Ever wondered how advanced aerospace engineering firms are tapping AI to reduce costs and lead times to launch successful flight missions?
For moon exploration, the Artemis program by NASA aims to land the first woman and next man on the Moon by 2024 using innovative tech to explore more of the lunar surface than ever before.
The firm behind this program, Lockheed Martin, has to continually improve the speed and efficiency of system diagnostics on its crafts to ensure viability and sustainability. Its executive VP Rick Ambrose said that this is done by “proactively analyzing telemetry data” with AI and machine learning “to streamline the work that employees do everyday.”
Across the space domain, the firm applies AI technology to provide proactive insights during design (of the Orion vehicle for the Artemis moon mission), production and for operational mission requirements.
Far out AI results
The Lockheed team has used AI to analyze unprecedented amounts of data produced by the Orion vehicle during testing. Within a four-hour period, its systems took data from nearly 150,000 sensors to establish over 22 billion logical relationships for analysis, and then used them to build a model of the vehicle’s normal operations.
The resulting models can be used to monitor all future tests of subsequent vehicles to compare expected and irregular behavior to analyze consistency and aid in regression analysis. Without these advanced AI and ML tools, it would be impossible for a single engineer to analyze massive amounts of data manually at the speed needed to analyze the data in its entirety.
Going forward, AI will continue to be applied on future missions in several ways, including future ground station support for customer satellite operations as well as expanding the application for human-rated systems to demonstrate an increase in speed of anomaly detection and root cause analysis on missions.
Ambrose noted: “The power of AI is leveraged across our entire enterprise, and we gain the resources to expand its abilities at scale across our internal operations,” alluding to the firm’s AI partner NEC. For several years, the two firms have been working together to evaluate the effectiveness of the latter’s System Invariant Analysis Technology (SIAT) for early production testing and operational scenarios.
As a result, Lockheed had integrated SIAT into the its Technology for Telemetry Analytics for Universal Artificial Intelligence (T-TAURI) AI platform. This has allowed the organization to drive proactive anomaly detection during the design, development, production and test phase of spacecraft development—even before applications in mission operations.
The SIAT advanced analytics engine uses sensor data to learn the behavior of systems, including computer systems, power plants, factories and buildings, so that it can automatically detect inconsistencies and prescribe resolutions. When integrated into Lockheed Martin’s T-TAURI platform, the system allows engineers to gain an exhaustive, holistic understanding of what they are developing, creating a foundational system for other advanced technologies like system-level digital twins.
Since the initial partnership between NEC and Lockheed Martin, the teams have already seen applications for early pathfinding to detect production anomalies and software regression testing. The partnership will be extended via a multi-year option, according to the latest announcements. Said NEC’s executive VP, CTO and Member of the Board: “We’re thrilled … that SIAT … provides value to Lockheed Martin as it addresses not only the development of complex systems, but also the uniqueness of deep space exploration. Together, we are committed to strengthening our solution and finding new ways to apply this technology.”
