NASA's exploration program envisions the utilization of a Deep Space Habitat (DSH) for human exploration of the space environment in the vicinity of Mars and/or asteroids. Communication latencies with ground control of as long as 20+ minutes make it imperative that DSH operations be highly autonomous, as any telemetry-based detection of a systems problem on Earth could well occur too late to assist the crew with the problem. A DSHbased development program has been initiated to develop and test the automation technologies necessary to support highly autonomous DSH operations. One such technology is a fault management tool to support performance monitoring of vehicle systems operations and to assist with real-time decision making in connection with operational anomalies and failures. Toward that end, we are developing Advanced Caution and Warning System (ACAWS), a tool that combines dynamic and interactive graphical representations of spacecraft systems, systems modeling, automated diagnostic analysis and root cause identification, system and mission impact assessment, and mitigation procedure identification to help spacecraft operators (both flight controllers and crew) understand and respond to anomalies more effectively. In this paper, we describe four major architecture elements of ACAWS: Anomaly Detection, Fault Isolation, System Effects Analysis, and Graphic User Interface (GUI), and how these elements work in concert with each other and with other tools to provide fault management support to both the controllers and crew. We then describe recent evaluations and tests of ACAWS on the DSH testbed. The results of these tests support the feasibility and strength of our approach to failure management automation and enhanced operational autonomy.