Skip to Content

UC04.3: Detailed indoor exploration with ANYmal (USAR)

Version 14.4 by Tjalling Haije on 2026/02/16 08:49
ObjectiveOB01: Increase effectiveness through exploration, sensing and communication means
OB04: Improve safety by providing overviews and warnings of environment conditions
Description

Enable safe, data‑driven interior reconnaissance and victim triage before human entry, with continuous FR health/location monitoring informing decisions, rotations, and evacuation.

TDP1a: Pattern: “Robots first, then humans.”, 2c: IDP "role-based, context-specific views", 
IDP 
ActorsTeam leader, Robot operator, Analyst, Safety officer (at HQ), Entry Team (FRs) (described at 4. Personas & Problem Scenarios and Direct Stakeholders)
Pre-Condition
  • Structure is flagged as unsafe for direct human entry; sector/ worksite allocated.
  • ANYmal is mission-ready (battery, controls, and payload checked).
  • FRs wearables distributed, connected, and calibrated.
  • C3I is online, role-based views are configurated (Operator/ Analyst/ Team leader/ Safety officer); 5G pods are installed for connection. 
Post-Condition
  •  
  • Health alerts handled; intervention performed if needed.
StatusHazards marked and/or victims detected; team proceeds with increased safety.

Action Sequence

  1. Deployment and Startup
     a. Team leader labels the buildings interior as robot-first (no human entry). 
     b. Robot Operator positions ANYmal at the threshold of the building while Analyst opens the control/analyst workspace of ANYmal in C3I.
     c. Wearables; Entry team confirms connection and baseline vitals. Safety officer sees green 'ready' status on the health/ location dashboard in C3I.
        d. Comms/ network connection check; confirms connection is stable for robot video/ LiDAR and wearable streams between worksite and HQ.
  2. Exploration and navigation
        a. ANYmal enters through the safest/ big enough opening with remote Operator tele-operating the tight passage.
        b. ANYmal builds a live 3D map (LiDAR) with the LiDAR and (depth) cameras while being tele-operated by the operator; Analyst interprets the camera feed and 3D map for victim and hazard detection.
        c. Real-time feed (video/ 3D map / thermal?) streams to C3I (to HQ); Analyst places markers on C3I (doors, unstable zones, possible victims).
  3. Victim detection and Robot/Operator control
        a. A candidate victim is detected on the LiDAR map/ camera view (with a confidence score?) by the Analyst; on request of the Analyst the robot Operator moves the ANYmal closer to verify.
        b. If equipped, ANYmal's arm can nudge light debris or open door to approach the possible victim detection for verification.
        c. Victim locations are confirmed, tagged, and prioritized in C3I with access contraints (blocked corridor, gas risk, etc.) for Team leader's plan.
  4. Team coordination and decision support
     a. An updated worksite map (victims, no-go areas, suggested routs?) is visable on the Team leader's tablet in C3I; and addional information is communicated with the Team leader by the Analyst.
     b. Team leader refines the plan (who enters, route, tools) and briefs Entry team. Decision rationale is logged in C3I for HQ. 
     c. ..
  5. Remote monitoring and escalation
     a. Part of the entry team is now inside of the building; while safety officer at HQ verfies continious FR vitals and positioning.
     b. If critical threshold is crossed (e.g. high heart rate/ temperature spike), Team leader recieves an on-dashboard critical alert while the affected FR (or/and buddy?) sees a local alarm (LED/ audio/ haptic).
     c. Escallation policy: alarms .

Open questions on technology capabilities

  1. Anymal:
    1. What sensors does the Anymal have? LiDAR, day-light camera, depth-camera, thermal camera, microphones, speakers? 
    2. How many of these sensors are available to the operator/analyst? 
    3. Is the operator UI connected with C3I?
    4. Can the operator share information via the operator UI connected to C3I to the command center? If so, in what form? Photo with image and text? A short video? A livestream? 
    5. Can the operator place a marker on the C3I map with a detection (e.g. hazard or victim)? Can information be attached to that marker? Can that marker be confirmed / prioritized / further specified with constraints (e.g. access constraints)? 
    6. Can the robot do object detection for finding victims? If so, how are results presented to the operator/analyst? 
  2. C3I:
    1. What information can be logged in C3I on the teamleaders tablet? Can suggested routes and no-go zones also be specified? 

Open questions on user needs

  1. HQ
    1. At what level of detail do they require Situation Awareness? Does this differ per situation? 
    2. Does HQ communicate always via the teamlead, or also directly with the robot operator/analyst? As in, does the teamleader "confirm" new markers etc in C3I before they become visible to HQ? 
    3. In what way does HQ want to maintain SA? Markers on C3I map with text description, with image and description, only image with category, multiple images, short video, livestream, live 3D map, live thermal camera stream, audio recordings.
    4. What are the tasks of HQ
    5. How many teams typically are coordinated by HQ? 
    6. How much time does HQ have available for maintaining SA and communicating with the teamlead etc on this part of the mission? 
  2. Teamlead
    1. At what level of detail does the teamlead require Situation Awareness? Which info do they need? Does this differ per situation? 
    2. In what way does teamlead want to maintain SA? Should they have enough info from inspecting the C3I map, do they require talking with the analyst, or do they require looking along to inspect the robot sensors? 
    3. What other tasks does the teamlead have? How much time do they have available for maintaining SA on the mission / talking with the analyst / talking with HQ / making plans, etc. 
    4. How many teams does the teamlead have? Are they always close to the operator/analyst (shouting distance), or not?
    5. How do the analyst and teamlead communicate? Via talking in person, or via radio? 
  3. Operator/Analyst
    1. What tasks does the Analyst have: analysing the robot sensor streams, instructing operator for where to go, updating the teamlead, updating the C3I interface, other tasks? 
    2. How does the Analyst communicate with others, via speech in person, via radio, via C3I? 

Table below need to be updated

Claims (title)FunctionEffect(s)Action Sequence Step(s)
CL1 Improve safety for respondersRobot enters unsafe or blocked zonesEFXX: Improves safety of responder by having robot explore unsafe or blocked zones instead 
→ Measured with FR path logs or near-incident reports.		2a
CL2 Improved FR SA SNAKE camera visual confirmationEFXX: Improves SA of FR by inspecting small spaces 
→ Measure with NASA TLX? 		4b
CL2 Improved victim detection rate in small spacesSNAKE camera visual confirmationEFXX: Victim discovery rate increased
→ Measured with number of victims detected & time to victim detection, compared to manual entry.		2b, 2c
CL3 Safer responder routingHazard identification via robotEFXX: First responder safety increased
→ Measured with map updates and avoided hazards.		2c, 3a
CL4 Trustworthy SAOperator-validated visualsEFXX: Trust in robot findings increased by operator confirmation
→ Measured with trust survey or follow-up interviews.		4a
CL5 Operational speed improvedRobot scouts ahead of teamEFXX: Operational speed improved
→ Measured with time-per-room metrics.		2a, 3c