Berkeley & Imperial

Workflow

From hazard findings to test evidence

We use hazard analysis to choose what to model and test. Findings from testing can update the model, the test scenarios and the hazard analysis.

  1. Hazard analysis

    Use failure modes and effects analysis (FMEA) and hazard and operability studies (HAZOP) to identify failure modes, deviations and safeguards. Turn priority findings into test scenarios.

  2. Modeling

    Represent the physical system, control logic and their interfaces. Define the operating modes, assumptions and limits that matter to the safety question.

  3. Simulation

    Establish how the system behaves in normal operation, during transitions and near its operating limits. Use the results to compare designs and define expected responses.

  4. Fault injection

    Introduce selected failures in the virtual test environment. Observe detection, alarms, interlocks and recovery, and compare the response with the expected behavior.

  5. Calibration

    Compare model responses with available test or operating measurements. Tune relevant parameters and document where the model can be relied on.

  6. Digital twin instantiation

    Configure the calibrated model for a specific asset or test setup. Connect it to the relevant hardware, control software or operating data for its intended use.

  7. Agentic AI

    Use AI agents to help prepare scenarios, run approved simulations and organize results. Engineers review the scenarios, interpret the evidence and make safety decisions.

Apply the workflow to your system

Tell us which failures, operating modes or design changes you need to examine.

Contact us