Skip to content

Sensing standoff and people-avoidance altitude are different numbers; no single UAV height is supported

Field Value
Record ID FINDING-0003
Type finding
Title Sensing standoff and people-avoidance altitude are different numbers; no single UAV height is supported
Status draft
Confidence medium
Updated 2026-09-09
Topics standoff_altitude, uav_operations, electromagnetic, gpr

Finding

Published sources do not support one UAV altitude that is simultaneously (a) high enough to stay clear of people, vehicles, and buildings in a cantonment and (b) low enough for utility GPR or conventional FDEM. Civil Part 107 sets a 400 ft AGL ceiling, not a survey height. Ground-coupled utility GPR is specified to remain within about two inches of the surface. FDEM papers put useful sensor height in the sub-meter to low-meter band, with safety already in tension at those heights. LCS-class EMI for long metallic conductors is the only reviewed modality with modeled detection remaining “easy” at 1–4 m AGL (Hartshorn STR models); that is not a flown LSL result and is not GPR. Dense/urban sites are described by the LCS authors as a poor UAS fit. In sparse layouts, finding two valves and drawing a straight line can locate a route; it does not identify lead.

Supporting Evidence

  • EVID-0017: §107.51(b) 400 ft AGL maximum; §107.39 restricts flight over people.
  • EVID-0015: GSSI ground-coupled rule ~1/10 wavelength / ~2 in for 270–900 MHz.
  • EVID-0014: FDEM desired 10–20 cm; pilots 50–150 cm; MiniExplorer optimum 50–130 cm; >130 cm failed their SNR rule; drone–sensor ≥150 cm.
  • EVID-0013: LCS 1/R² vs compact 1/R⁶; modeled 1, 2, 4 m AGL >20 dB STR for a 1 m deep long wire; 8 m weaker; speculated 1.5 m hover on flat terrain; urban/wooded called poor UAS fit; pipe data were hand-carried.
  • EVID-0016 (preprint): example 5 m AGL FDEM flight over an inferred metallic pipeline; not peer-reviewed; not LSL.
  • EVID-0003: ULEMA/LCS aerial mounting is a capability claim, not an Edison UAV demo.
  • EVID-0020 / EVID-0021: JIFX ACP and NPSINST 3700.1A — R-2504, 300-ft buffer, no McMillan cantonment overflight; sparse/unpopulated terrain.
  • EVID-0022: UAV air-launched GPR at 2.3 m AGL for landmine-like targets; ground-coupled has better SCR.
  • EVID-0026: 1000 MHz GPR raised 0–0.35 m; field diffractions only visible at 0.05 m AGL (preprint).
  • EVID-0027: UAV FDEM height 0–7 m for soil ECa, not pipes.

Evidence Against / Qualifications

  • LCS modeled heights are for a 100 m wire, not a small-diameter lead service line.
  • Karaoulis and Hartshorn envelopes differ (0.5–1.3 m vs 1–4 m) because the sensors differ.
  • Vendor GPR gap rule is not a UAV experiment.
  • DoD installations may be stricter than Part 107. JIFX/Camp Roberts ACP is now sourced (EVID-0020, EVID-0021): R-2504 range airspace, 300-ft interior buffer, no McMillan cantonment overflight. That answers conflict-avoidance geometry for the named demo site; it still does not pick a sensing height.

Confidence Assessment

High that 400 ft is not a sensing altitude. Medium that GPR UAV at cantonment-safe height is a different (worse-coupled) regime than cart GPR. Medium that LCS-class EMI is the only reviewed path with modeled meter-scale AGL. Low for any numeric Pd or localization loss vs ground.

Remaining Unknowns

  • Measured GPR SNR vs height for service-line diameters in Edison soils.
  • Measured ULEMA/LCS STR vs height on lead/copper SLs (Task 2).
  • UAV motion penalty vs gantry at the same height (Task 3).
  • Installation-specific airspace at sites other than Camp Roberts/JIFX (NAVFAC facilities still unsourced).
  • Pipe-specific EMI height-response curves (Cheng 2025 sweeps height for soil ECa, not pipes; Hartshorn remains modeled long-wire).
  • Measured GPR SNR vs height on service-line diameters (Booth/García-Fernández are elevated GPR but not LSL pipes).

Operational Implication

Treat controlled elevated collection and UAV collection as separate. Do not advertise a universal airborne survey. Cantonment: terrestrial or very constrained UAV. Open corridor/range: possible UAV niche for long-conductor EMI mapping, not for GPR-equivalent coupling.

Experimental Implication

Task 2 should sweep sensor height at Edison (at least the Hartshorn 1–4 m LCS band and a near-ground GPR/FDEM band) without a UAV. Task 3 only after that.

Proposal Implication

Answer the reviewer: height to avoid people is a legal/safety floor that conflicts with GPR/FDEM coupling; the project will measure the trade, not claim a number. Sparse-area straight-line inference is a routing prior, not material ID.

Linked Evidence

  • EVID-0003
  • EVID-0013
  • EVID-0014
  • EVID-0015
  • EVID-0016
  • EVID-0017
  • EVID-0020
  • EVID-0021
  • EVID-0022
  • EVID-0026
  • EVID-0027