BEAM is a method of real-time, automated analysis and diagnosis applicable to a broad class of complex electromechanical systems, including spacecraft, aircraft, and process-control systems. Some aspects of its operation were described in “Reusable Software for Autonomous Diagnosis of Complex Systems” (NPO-20803), NASA Tech Briefs, Vol. 26, No. 3 (March 2002), page 33. Presented here is an expanded overview of the method, outlining its components and their function.

BEAM was conceived to accelerate diagnosis and to relieve human operators and ground control computers of the burden of diagnostic data collection and analysis. This is performed through a real-time fusion and analysis of system observables, including not only performance and sensor data but also knowledge of executing software and commands sent to the system. In the case of a spacecraft, BEAM would enable onboard identification of anomalous conditions, thereby obviating the need to telemeter large quantities of sensor information.

The BEAM formalism is based upon a reduction of the observables in a complex physical system to a compact set of coupled critical observables, which are tracked to analyze the state (or “health”) of the system in real time. The coupled observables determine the information space of the physical system, and monitoring its invariants allows all events, responses, deter- ioration, anomalies, and failures to be detected and isolated with high precision. In contrast to classical model-based approaches in which one either (1) attempts to explicitly compare observed data to model predictions, or (2) relies upon a coarse operating envelope in the form of redlines, BEAM is highly adaptable and sensitive to complex nonlinear behaviors indicative of such faults.

The mathematical foundation of BEAM includes the following building blocks:

  • The System Invariance Estimator (SIE) automatically constructs fundamental information invariants from multichannel data. Comparison of the invariants allows system diagnosis, identifies which observables are significant, and quantifies deviation and dependency between events.
  • The Channel Coupling Operator (CCO) provides an embedded, algorithmically constructed means of relating sensed transitions and commands. This permits oversight of software executing in conjunction with the sensed hardware.
  • The Data Fusion Operator (DFO) determines the proper combinations of observables to provide an instantaneous estimate of the system information state. This estimate serves as a single, event-based, health metric.
  • The Variable-Fidelity Discontinuity Operator (VDO) utilizes an adaptive wavelet transformation to detect and amplify the onset of transitions or incipient faults. This operator is a highly sensitive means of performing singlechannel analysis that is adaptable to very short or very long event periods.
  • The Operating Map and Back-Projection Operator (BPO) represent the downlink and reconstruction elements of BEAM. The first is a compact parameter set based upon the other modules, which efficiently encapsulates specific features of fault events. This information is expanded by the BPO to allow causal reconstruction of the anomaly track.

These modules are applicable to nearly any system and can be trained using brief examples or approximations of nominal operating data. In combination these modules permit system analysis at the local and global level, and are responsive to hard faults, incipient faults, and anomalies that lie beyond the training envelope. BEAM is intended to reduce or eliminate the need for external diagnosis but also provide capability for autonomous detection and isolation of faults and degradation.

This work was done by Sandeep Gulati and Ryan Mackey of Caltech for NASA’s Jet Propulsion Laboratory.

In accordance with Public Law 96-517, the contractor has elected to retain title to this invention. Inquiries concerning rights for its commercial use should be addressed to

Intellectual Property group
JPL
Mail Stop 202-233
4800 Oak Grove Drive
Pasadena, CA 91109
(818) 354-2240

Refer to NPO-20827, volume and number of this NASA Tech Briefs issue, and the page number.

Topics:
Aircraft Computer software and hardware Data management Education and training Information Technology People and personalities Sensors and actuators Spacecraft Technical review Vehicles
This Brief includes a Technical Support Package (TSP).
Document cover
Beacon-Based Exception Analysis for Multimissions

(reference NPO-20827) is currently available for download from the TSP library.

Don't have an account?

Magazine cover
NASA Tech Briefs Magazine

This article first appeared in the September, 2002 issue of NASA Tech Briefs Magazine (Vol. 26 No. 9).

Read more articles from the archives here.

Overview

The document presents an overview of the BEAM (Beacon-Based Exception Analysis for Multimissions) technology developed by the Jet Propulsion Laboratory (JPL) under NASA's sponsorship. BEAM is designed to enhance the diagnosis and monitoring of complex systems, particularly in aerospace applications, by automating the collection and analysis of diagnostic data. This technology aims to alleviate the burden on human operators and ground control systems by enabling real-time identification of anomalies onboard spacecraft, thus minimizing the need for extensive telemetry data transmission.

The BEAM methodology is based on a mathematical framework that reduces complex physical systems to a compact set of coupled critical observables. These observables are tracked to analyze the system's health in real time, allowing for the detection and isolation of events, responses, and failures with high precision. Unlike traditional model-based approaches, which either compare observed data to model predictions or rely on coarse operating envelopes, BEAM is adaptable and sensitive to complex nonlinear behaviors indicative of faults.

Key components of the BEAM framework include:

  1. System Invariance Estimator (SIE): Constructs fundamental information invariants from multichannel data, identifying significant observables and quantifying dependencies between events.
  2. Channel Coupling Operator (CCO): Monitors the relationships between different observables.
  3. Data Fusion Operator (DFO): Provides an instantaneous estimate of the system's information state through nonlinear combinations of observables.
  4. Variable-Fidelity Discontinuity Operator (VDO): Detects and amplifies transitions or misbehavior in the system.
  5. Operating Map: A compact representation of system health and state.
  6. Back-Projection Operator (BPO): Reconstructs the causal track of anomalies.

The development of BEAM addresses several challenges in aerospace operations, including the elimination of backup telemetry systems and the need for a robust health and state summarization technology. By integrating these capabilities, BEAM supports JPL's vision of efficient beacon-based operations, ultimately reducing operational costs and improving system reliability.

In summary, BEAM represents a significant advancement in autonomous diagnosis for complex systems, providing a comprehensive solution for real-time health monitoring and anomaly detection, which is crucial for the future of aerospace missions.