The ISStrider project is developing model-based reasoning, visualization and document retrieval capabilities for mission and spacecraft operations.
The goals of the ISStrider system are to decrease mission operations costs and increase safety margins through the use of model-based reasoning methods. Real-time problem diagnosis and recovery will eventually pave the way to tightly integrated model-based mission operations, from planning to post-flight inspections.
ISStrider is being applied to the International Space Station (ISS) but is relevant to the development of sustainable Project Constellation missions, assisting with the following objectives of that program:
(i) affordability -- reduce ground/onboard time in determining root-cause and data analysis,
(ii) reliability/safety -- make the onboard systems locally self-reliant to make missions as safe as reasonably achievable (ASARA),
(iii) effectiveness -- facilitate timely and accurate recovery from on-board failures,
(iv) flexibility -- ISStrider will be able to incorporate subsystems as they become available.
ISStrider consists of three capabilities:
These three capabilities are combined into the Advanced Diagnostic System (ADS), which utilizes the Diagnostic Data Server (DDS) to subscribe to ISS telemetry as well as ISS training and simulation scenarios, and incorporates the Netmark document server to subscribe to ISS documents.
Caution and Warning Fusion (CWF) -- models the subsystems of the spacecraft and their interactions in order to provide a capability for enforcing global consistency between individual Caution and Warning events. Currently, mission controllers and astronauts are instructed to resolve each CW event in chronological order. However, if the CW events are related, than an integrated novel recovery may be the correct response. To determine common root-causes between CW events and to suggest appropriate recovery actions, we are developing the CWF capability based upon a model-based diagnosis and recovery inference engine. Currently we utilize the L2 inference engine, whose predecessor Livingstone flew on the Deep Space One spacecraft. Additional diagnosis engines can be used as well.
Caution and Warning Cube (CW^3) provides a set of components which allow users the ability to rapidly visualize and manipulate raw data, as well as the results, from the model-based reasoning tools. Each visual component serves as an entry point into a network of related displays and information. Types of displays include: (i) the time series of ISS Caution and Warning events and diagnoses (ii) spatial location of parameters in schematics (iii) explanations of root-cause(s) and (iv) presentation of recovery action(s).
Real-time Knowledge Management (RKM) alleviates the challenge of finding the relevant documentation required to analyze ISS anomalies by indexing ISS documents to the L2 models of the ISS subsystems. When the CWF system determines the root-cause(s), the relevant documentation is automatically presented to the mission controllers and the crew. This capability is being developed in cooperation with Netmark (NASA Ames). The planned types of documents indexed include CW fault tree logic, Architectural Design Documents (ADD), Interface Control Documents (ICD), source code, schematics, and recovery procedure diagrams.
The International Space Station (ISS) is one of the most complex aerospace vehicles ever constructed. It is a unique spacecraft being developed using a staged approach. The presence of several subsystems, some fully developed and others in the process of development, and their interaction with the control software (that is also being developed in a staged manner), can lead to unanticipated problems.
The current on-board diagnostic system consists of a large set (~10K) of caution and warning (CW) events that fire in response to anomalies. In some cases, CW events fire en masse as a result of cascading effects and mission controllers and astronauts must determine root causes in order to resolve the anomalies.
With over 10 interacting subsystems, controlled by 1 million lines of Ada code executing on over 30 computers, ISS requires a very large team of mission controllers to analyze and resolve anomalies. As such it serves as an excellent test-bed for the development of model-based mission operations tools which explore methods to increase safety in a sustainable manner.