Go to the Indico Home Page Oldest event Older event Up to category Newer event Newest event Printable version
  • Compact style
  • Indico style
  • Indico style - inline minutes
  • Indico style - numbered
  • Indico style - numbered + minutes
  • Indico Weeks View
    • Switch to the management area of this event
    Login

    Robotics for Access and Patrols

    → Europe/Zurich
    CERN

    CERN
    Hide

    Meeting Notes

    Topic: EN-AA Robotics Use Cases for FCC Concept Design
    Focus: Robotic applications to improve Safety, Reliability, and Availability of the FCC access, patrol, alarm, fire, and gas systems.
    Date: Based on transcript
    Participants (identified): Scott Williams, Tomasz Ladzinski, Nicolas Venzin, Michael Dole, Virama Ekanayaka, Morgan Nightingale and Hannes Gamper

    Meeting Objective

    The meeting aimed to identify and refine robotic use cases for EN-AA within the FCC concept design, focusing on:

    • Patrol and access safety functions.
    • Fire and gas detection applications.
    • Safety validation considerations.
    • Requirements and operational needs that robotics systems must satisfy.
    • Potential collaboration opportunities between EN-AA and the FCC Robotics team.

    Discussion Summary

    1. Robotic Patrols for Personnel Safety

    The primary discussion focused on replacing or significantly reducing human patrols with robotic patrol systems. Potential environments include:

    • Main tunnels, using RMIS for patrols
    • Alcoves, where the robot must inspect at multiple heights and move between several inspection points
    • Service caverns, including areas with controlled-access doors

    Currently, personnel begin patrols at the deepest point of an area and work their way back towards the exit. At the LHC, a patrol can take several hours per site.

    Two types of patrol are performed:

    1. Systematic patrols when closing the machine after a Long Shutdown
    2. Local patrols following a patrol breach; these occur relatively frequently

    Current Situation

    Today, patrols are performed by teams of people who:

    • Traverse accelerator zones before beam operation.
    • Verify that no personnel remain underground.
    • Sequentially validate sectors using patrol key stations.
    • Close and secure sectorization doors.
    • Confirm patrol completion before beam authorization.

     

    Patrol personnel carry specific keys. If somebody enters and exits correctly using the designated keys, the patrol status is retained, provided that the large access doors are not used.

    A lost patrol prevents beam operation or can stop operation if patrol integrity is compromised.

    FCC Challenge

    Participants agreed that performing patrols using humans at FCC scale is unrealistic because:

    • The machine is significantly larger than existing CERN accelerators.
    • Patrol durations could be several hours.
    • Staffing requirements would be excessive.
    • Human attention decreases during long repetitive inspections.

    The consensus was that robotic patrols are effectively required for FCC.

    2. Patrol Robot Functional Requirements

    Environment Coverage

    The robot must be capable of:

    • Inspecting at multiple heights.
    • Looking above cable trays and below pipework.
    • Navigating cluttered environments.
    • Accessing alcoves and service caverns.

    Mobility

    For alcoves and caverns, robots may need:

    • Stair-climbing capability.
    • Autonomous navigation.
    • Reliable localization.

    Potential technologies mentioned:

    • SLAM-based navigation.
    • LiDAR.
    • Radar.
    • Visual odometry.
    • Fixed markers.

    Door and Sectorization Handling

    A major challenge discussed was sectorized areas.

    Robots must:

    • Pass between sectors.
    • Maintain patrol integrity.
    • Potentially open and close doors.
    • Verify doors are closed.
    • Interact with patrol validation points.

    Alternative concepts include:

    • Automated hatches near the tunnel ceiling for rail robots.
    • Robot-specific passageways that avoid reopening patrol sectors.

    This was considered an important requirement for maintaining availability.

    3. Human Detection Requirements

    Participants emphasized that simply detecting a standing person is insufficient.

    The patrol system must detect:

    • People lying on the floor.
    • Partially visible persons.
    • Unconscious persons.
    • Limbs or body parts visible behind equipment.

    Detection Technologies Discussed

    Potential sensing approaches include:

    • Standard cameras.
    • AI-based object detection.
    • Thermal imaging.
    • Radar-based heartbeat detection.
    • Respiration monitoring.
    • Optical pulse detection.

    The robotics team indicated that several of these technologies have already been tested.

    4. Patrol Validation and Safety Integrity

    A key topic was how to demonstrate that a robotic patrol is sufficiently reliable to replace human patrols.

    Discussion Points

    The system must provide:

    • Evidence that all areas have been inspected.
    • A verifiable patrol completion process.
    • Sufficient risk reduction to satisfy CERN safety requirements.

    Additional independent safety barriers were discussed:

    • Wearable devices.
    • Access control records.
    • Localization systems.
    • Human confirmation procedures.

    The group agreed that robot patrols will require a formal safety validation methodology.

    5. Localization and Communication

    Several requirements emerged:

    Localization

    The system should provide:

    • Real-time localization.
    • Position awareness in caverns and alcoves.
    • Ability to identify sectors containing detected personnel.

    Very high precision is not currently required for patrol purposes; sector-level localization may be sufficient.

    Communication

    Robots should support:

    • Real-time communications.
    • Two-way voice communications.
    • Safety Control Centre interaction.

    This is particularly important if a person is located during a patrol.

    6. Wearable Technology

    Wearables were discussed as an additional safety layer.

    Potential functions include:

    • Tracking worker locations.
    • Providing evacuation guidance.
    • Supporting emergency response.
    • Offering an independent check against patrol results.

    However:

    • Wearables cannot replace patrol systems.
    • Users may remove or forget devices.
    • They are viewed as an additional risk-reduction layer rather than a primary safety system.

    Fire and Gas Detection Discussion

    Robotic Fire Detection

    The second major topic was the use of robots for fire detection and response.

    Proposed Functions

    Robots could:

    • Detect smoke.
    • Detect flames.
    • Detect temperature increases.
    • Confirm alarms.
    • Localize fires.

    This could reduce false callouts and improve emergency response.

    Localization Requirement

    Michael Dole indicated that:

    • Precise localization is unnecessary.
    • Localization within approximately 10-50 meters would likely be sufficient.

    Fire Suppression

    An additional use case discussed was:

    • Early-stage fire suppression.

    Existing FCC emergency concepts already include:

    • Fire detection.
    • Initial response capability.

    Current high-level FCC emergency response requirements include:

    • Ability to reach any tunnel location within 10 minutes.
    • Ability to respond to a fire of up to approximately 5 MW during the early fire growth phase.

    Continuous Monitoring

    The possibility of continuous robotic monitoring was discussed.

    Rather than relying solely on fixed detectors:

    • Robots could continuously patrol.
    • Robots could provide situational awareness.
    • Robots could monitor environmental conditions.

    This appeared feasible but would complement rather than replace fixed fire protection systems.

    Additional Robotics Opportunities Identified

    Fire Protection Maintenance

    Robots could automate annual testing and inspection of:

    • Smoke detectors.
    • Thermal sensors.
    • Fire detection devices.

    Potential benefits:

    • Faster execution.
    • Reduced manpower requirements.
    • Night-time operation.

    Alarm System Testing

    Robots could support testing of:

    • Evacuation alarms.
    • Loudspeaker systems.

    Given the expected FCC scale:

    • Thousands of loudspeakers may require periodic validation.
    • Robotic testing could significantly reduce labour requirements.

    Companion Safety Robots

    An additional future concept discussed was using robots as worksite companions.

    Potential functions:

    • Local hazard monitoring.
    • Fire and gas monitoring.
    • Personnel observation.
    • Emergency communications.
    • Remote supervision of work activities.

    Participants felt this could provide significant safety benefits but would not replace fixed safety systems.

    Reliability and Availability Benefits Identified

    Safety

    Robots could:

    • Remove personnel from hazardous patrolling activities.
    • Improve personnel detection capability.
    • Support emergency intervention.
    • Improve fire detection and localization.
    • Provide additional safety barriers through monitoring and communications.

    Reliability

    Robots could:

    • Perform repeatable inspections.
    • Provide continuous monitoring.
    • Reduce dependency on human attention.
    • Improve consistency of patrol execution.

    Availability

    Robots could:

    • Reduce patrol times.
    • Reduce recovery time after patrol loss.
    • Re-patrol only affected sectors.
    • Reduce downtime associated with access and safety procedures.
    • Accelerate inspection and testing campaigns.

    Action Items

    Action Owner
    Prepare and distribute meeting notes and extracted requirements. Scott Williams
    Draft a structured list of robotic patrol needs and justifications. Scott Williams
    Share FCC robotics requirement documents (operational safety and emergency response requirements) with EN-AA. Robotics Team
    Review robotics requirements and provide comments from fire safety perspective. Michael Dole / EN-AA
    Develop fire detection and localization functional needs for robotic systems. EN-AA
    Explore integration of EN-AA fire detection research into future robot concepts. Michael Dole & Robotics Team
    Organize a follow-up meeting involving the new EN-AA graduate student and robotics team. Scott Williams / Michael Dole
    Define software and hardware interface requirements for future sensor integration. Robotics Team & EN-AA
    Provide preliminary cost estimate for SES robot prototype. Robotics Team
    Provide indicative SES prototype development schedule. Robotics Team
    Investigate funding opportunities, potentially shared across multiple FCC stakeholders. EN-AA & Robotics Stakeholders
    Review which robotic functions must be treated as safety-critical systems. EN-AA
    Evaluate patrol validation methodology and safety certification approach. EN-AA & Robotics Team
    Begin testing human-detection algorithms in FCC mock-up facilities. Robotics Team

    Key Takeaways

    1. Robotic patrols are considered essential for FCC, primarily due to scale and manpower limitations.

    2. The most critical patrol capability is reliable human detection, including unconscious, hidden, or partially visible persons.

    3. Sectorization and door management are major design challenges and will strongly influence robot architecture.

    4. Robust localization and two-way communications are key requirements for safety operations.

    5. Wearables and access control systems are valuable independent safety barriers, but cannot replace robot patrols.

    6. Fire safety presents multiple robotics opportunities:

      • Fire detection
      • Alarm confirmation
      • Fire localization
      • Early fire suppression
      • Continuous environmental monitoring

    7. Significant additional value exists in using robots for inspection, testing, and maintenance of EN-AA systems.

    8. A major outstanding challenge is functional safety validation, particularly for AI-based human detection algorithms.

    9. There is strong interest in collaboration between EN-AA and the FCC Robotics team, particularly around a future SES robot prototype and integrated fire-detection technologies.

    10. The next phase is to convert these operational needs into structured requirements and assess which functions must satisfy formal safety-critical certification requirements.

    There are minutes attached to this event. Show them.
    The agenda of this meeting is empty