===== Automation (AFIRA: Automated Fault Identification and Recovery Agent) =====

To support fault management, TeamCAMS includes the **Automated Fault Identification and Recovery Agent (AFIRA)**. AFIRA provides different levels of assistance, ranging from no support at all to fully automated fault handling.

The following levels of automation are available:

  * **LOA 1 - No Assistance**
    * AFIRA is disabled.
    * Operators must detect, diagnose, and resolve faults entirely on their own.

  * **LOA 2 - Fault Detection**
    * AFIRA detects the occurrence of a fault.
    * Visual and audible alerts are provided.
    * Diagnosis and corrective actions remain the responsibility of the operator.

  * **LOA 3 - Fault Diagnosis**
    * AFIRA detects the fault and identifies its likely cause.
    * The operator remains responsible for implementing corrective actions and requesting repairs.

  * **LOA 4 - Procedure Guidance**
    * AFIRA provides both a diagnosis and a recommended sequence of corrective actions.
    * The operator decides whether and how to follow the proposed procedure.

  * **LOA 5 - Assisted Recovery**
    * AFIRA proposes both diagnosis and corrective actions.
    * Actions are carried out automatically after operator approval.

  * **LOA 6 - Automatic Recovery**
    * AFIRA automatically diagnoses and resolves faults.
    * Corrective actions are executed automatically unless explicitly vetoed by the operator within a predefined time window.

===== Adaptive and Fallible Automation =====

In addition to fixed automation levels, TeamCAMS supports more advanced automation concepts for research purposes.

When **adaptive automation** is enabled, AFIRA can recommend increasing or decreasing the level of automation based on the operator's performance during the session. This allows researchers to study how dynamic allocation of responsibility influences workload, performance, and trust.

AFIRA can also be configured as **fallible automation**. In this mode, the system may occasionally provide incorrect assistance, such as reporting a fault that does not exist or proposing an incorrect diagnosis. This functionality allows researchers to investigate important topics such as automation reliability, trust in intelligent systems, and human responses to automation errors.

Together, faults, repairs, and AFIRA provide a realistic environment for studying fault diagnosis, decision making, teamwork, human-automation interaction, and adaptive support strategies in complex operational systems.