Conceptual engineering study. AHV-NDR is a simulation-only capstone concept — no vehicle exists and no physical or field testing has been performed. All figures below are model outputs, not measured data.
AHV-NDR · Conceptual Study — Autonomous Disaster Recovery System
An autonomous heavy vehicle for nuclear disaster recovery.
A single-machine, single-battery mission profile inside a nuclear exclusion zone, and the energy analysis behind why it's feasible on one charge.
01 · Problem
Radiation dose, not physical barriers, defines the exclusion zone.
1986, Chernobyl. 2011, Fukushima. When disaster strikes a nuclear site, recovery work is needed for hours. Human occupational dose limits allow only minutes.
HOW LONG A HUMAN OPERATOR CAN WORK (radiation dose limit)
45 min
HOW LONG THE MISSION TAKES
8 hours
Every minute inside the zone costs a worker part of their lifetime dose allowance. Sustained work — clearing debris, delivering supplies, recovering hazardous objects — is not viable for human crews at these dose rates.
02 · System Overview
The AHV-NDR concept.
40 tonnes, fully electric, designed for unattended operation inside the exclusion zone. Select a hotspot below to see what each subsystem does and what fails without it.
Illustrative, not CAD-accurate. Roughly the footprint of two parked school buses.
SELECT A SYSTEM
Eight systems, one purpose
Every part of the AHV-NDR answers one question: how do you do eight hours of heavy work in a place where nothing can be refuelled, repaired, or rescued?
40 tmass
8 × 4 mfootprint
750 Vsystem voltage
387 kWhbattery
0 Lfuel required
03 · Mission Profile
Four tasks, one battery, no resupply.
In a single deployment, the AHV-NDR is modelled performing four tasks in sequence. Click each card for detail.
📦
PHASE 1
Payload Delivery
Proof of manipulation
Supplies and monitoring equipment carried into the zone and placed at a specified location. Validates manipulation accuracy before the higher-load tasks.
🪨
PHASE 2
Debris Clearance
Proof of power
Collapsed structures block the route and must be cleared by digging and dragging. This is the highest-draw phase in the model: hydraulic work consumes battery faster than driving.
🎯
PHASE 3
Object Recovery
Proof of precision
A contaminated object is lifted, contained, and carried out. Requires precise remote control from an operator kilometres away, over a wireless link with round-trip latency.
🏠
PHASE 4
Return to Base
Non-negotiable
A stranded vehicle becomes 40 tonnes of contaminated debris that would itself need recovery. The return leg is reserved energy, ring-fenced from the moment the mission starts.
04 · Energy Budget
Battery capacity is the binding constraint.
Every task costs energy and there is no recharging or refuelling inside the zone, so the mission plan is an allocation of a fixed 387 kWh budget across tasks.
A 5% margin is thin. The operator has to choose: attempt all three tasks in one trip, or drop a task and return with more reserve? Toggle the tasks below to see the effect on margin:
Safety margin: 5%— tight; little margin for unplanned contingencies.
05 · Cross-Validation
Two independent models, cross-checked.
The mission energy estimate was built two different ways. Disagreement between them would indicate a modelling error; agreement is a (necessary, not sufficient) sanity check.
Model A
First-Principles Physics
Built from the ground up: rolling resistance, grade climbing, hydraulic work, drivetrain losses. Pure equations — the kind you can check by hand.
=THEY MATCH
Model B
Industry Simulink Model
The same mission simulated in MATLAB Simulink — the tool used to validate real electric vehicles, aircraft, and spacecraft.
1.0×
CROSS-VALIDATION RATIO — INDEPENDENT MODELS, IDENTICAL ANSWER
Both models converge on the same result: the modelled mission is feasible on a single 387 kWh charge, with the return leg reserved. This is a model-based conclusion, not a field-tested one.
Modelling note: the first version of the first-principles model disagreed with the Simulink model by nearly 7×. The discrepancy was traced to underestimated hydraulic loads and missing temperature effects; the model was rebuilt with a proper equivalent-circuit battery representation and recalibrated against published excavator field data. After the fix, both models agree to within rounding. It's included here because the resolution process is more informative than the final agreement alone.
06 · Human-Machine Split
Autonomous execution, human command.
The vehicle navigates and executes autonomously, but a human operator commands the mission remotely, connected by a wireless link.
● CONTROL ROOM — SAFE ZONE
The human
Sees through the vehicle's sensors
Approves each mission phase
Watches the battery like a hawk
Makes the judgment calls
WIRELESS LINK · ~200ms
● EXCLUSION ZONE
The machine
Executes with machine precision
Navigates obstacles in real time
Reports everything it senses
Knows its own limits
When the vehicle encounters an obstacle not present in the satellite survey, it must navigate around it autonomously; the operator can review the decision after the fact but the ~200 ms round-trip latency rules out real-time manual control for obstacle avoidance.
07 · Failure Modes & Mitigations
Margins, redundancy, and failure protocols.
Three failure scenarios considered in the design, and the mitigation modelled for each. Click to expand.
MITIGATION
The return leg is reserved from the working budget before the mission starts. The energy models are cross-validated (see Section 05), and the mission plan carries a safety margin on top of that reserve.
MITIGATION
Real-time LiDAR and thermal sensors rebuild the map as the vehicle moves. Autonomous navigation reroutes around surprises, and the operator can override at any moment. The plan is a starting point, not a promise.
MITIGATION
The vehicle carries a 30-second autonomous buffer — enough to finish its current motion safely and stop. If the link doesn't recover, it falls back to a pre-planned return-to-base protocol. Losing contact never means losing the vehicle.
08 · Rationale
Why this is feasible now.
THE OLD WAY
Manned excavators, rotating crews
45-minute work windows per person
Human dose risk on every shift
Evacuation delays measured in days
Work stops when limits are reached
THE AHV-NDR WAY
Autonomous platform, remote command
8-hour continuous work windows
Zero humans inside the zone
Deployment within hours
Work stops when the job is done
Nuclear disaster recovery has historically been limited by human dose tolerance. Improvements in electric drivetrains, battery energy density, and autonomy make an unattended alternative technically plausible for this class of mission.
09 · Future Extensions
Directions the concept could extend to.
Not modelled in this study; listed as plausible follow-on directions.
🤖Multi-vehicle coordinationMultiple units sharing one mission plan and one collective energy budget.