Integrating PSR Traverse Plans With Sound-Derived Geometry
The Problem
A Volatiles Investigating Polar Exploration Rover concept simulation at a lunar-analog test site near Mauna Kea hit a wall at the edge of a simulated permanently shadowed region. The route planner depended on visual terrain classification up to the shadow line, then fell back to low-confidence hazard maps inside the PSR. When the rover sim was told to enter the dark zone on a battery-only excursion, the planner refused to commit to a specific path because it had no geometry information beyond the shadow boundary. Meanwhile, passive acoustic recordings from a scouting traverse had captured echo-geometry data for roughly 40 meters past the shadow line, but no interface existed to feed that geometry into the traverse planner.
This is the planning gap for PSR operations. VIPER's mission profile targeted roughly 37 km over 100 Earth days with 27 waypoint stations and 4 PSR excursions, all on battery-only power inside shadows. Traverse planners for PSRs cannot use the standard orbital imagery-driven workflows because cameras do not see into shadow. IOPscience's VIPER mobility paper documents how mobility models have to compensate for missing slope and roughness information on unshadowed approaches, and the gap only widens inside the PSR. Without additional priors, the planner defaults to conservative routes that waste PSR dwell time on safety margins.
Artemis crewed operations face the same question at a different scale. LPI analysis of Artemis III EVA capability shows astronauts reaching PSRs within roughly 2 km of landing sites, but traverses need independent geometry priors to route safely. ScienceDirect's automated traverse work for lunar south pole PSRs minimizes metabolic workload against known terrain, but when terrain is unknown inside shadow, the optimization has no ground to stand on.
The PSR planning gap is also where rover adaptation becomes critical, because acoustic data captured during PSR scouting routinely reveals new voids that were not visible in pre-traverse imagery. A planner that cannot fold those discoveries into the active traverse forces the rover to commit to a path that may have become suboptimal as new geometry emerges, which is exactly the scenario where mid-traverse adaptation pays off.
The thermal envelope inside PSRs makes the planning gap even sharper. Permanently shadowed regions sit below 110 K continuously, which means rover-mounted lithium batteries lose cycle capacity and EVA-suit thermal control has to compensate for sustained heat loss to the regolith. A traverse plan that treats shadow entry as a 90-minute bounded excursion is making implicit thermal assumptions that depend entirely on whether the planned path is actually traversable as planned. If the rover stops mid-PSR to re-evaluate a hazard, battery capacity drops faster than the recovery margin, and the excursion converts from a science success into a survival exercise. VIPER's mission concept explicitly excluded extended dwell inside shadow because the thermal margins did not close on multi-hour stops, so any acoustic-prior method that lets a traverse commit to a path without mid-shadow re-evaluation directly extends the science envelope rather than just optimizing routing.
The Solution: Acoustic-Prior PSR Routing
EchoQuilt's PSR integration layer lets route planners treat passive acoustic geometry as a prior on the same footing as orbital DEMs and hazard maps. A scout rover or EVA-carried sensor captures passive recordings during approach and brief intrusions into the shadow zone. Those recordings stitch into the quilt and produce a geometry patch with explicit confidence bounds. The route planner then consumes the quilt as a terrain layer, using high-confidence patches as hard constraints and low-confidence patches as soft priors.
This lets PSR excursions commit to specific paths rather than conservative safety boxes. A battery-only excursion with a 90-minute window inside shadow no longer has to spend 30 minutes re-scanning each step. Instead, it runs the pre-planned traverse, captures additional acoustic data to extend the quilt deeper into the PSR for the next excursion, and exits with battery margin intact. The pattern matches the safe mission-level path planning approach described in arXiv's safe PSR path planning work, where formal methods produce battery-feasible routes against precomputed terrain priors.
Crucially, the acoustic quilt also captures skylight and pit-rim geometry that orbital data can miss. Science's LRO LEND hydrogen mapping identified water-ice-relevant PSRs but did not resolve pit-rim geometry at the meter scale. The acoustic quilt provides exactly that scale, which matters because PSR entries often route around pit rims, and centimeter-level rim variation determines whether a rover can roll or has to detour. Integrating the quilt into the planner means each planned waypoint carries a reason: pass here because the acoustic patch shows a stable rim, avoid there because the patch shows a void below.
For crewed traverses, the same integration drops astronaut metabolic risk. If a traverse segment can be routed across an acoustically confirmed flat regolith patch rather than a visually ambiguous shadow zone, metabolic workload optimization in ScienceDirect's automated traverse work can be applied meaningfully. Without the prior, the optimizer has to assume worst-case terrain, which inflates workload estimates and shortens EVA time.
PSR integration connects directly to ISRU traverses because water-ice extraction sites are often inside PSRs and need confirmed route geometry before any equipment can commit to entry. The acoustic quilt is what gives the ISRU planner a route to score, rather than just a bounded shadow box that may or may not contain a path.
Advanced Tactics
Capture two scouting passes per PSR entry: one along the intended route and one parallel to it at roughly 3-5 meters offset. The parallel pass produces enough cross-track information for the quilt to resolve local slope, not just path centerline geometry, which matters when the planner needs to choose between two near-parallel paths. The cost is minor additional power; the payoff is a route choice the planner can defend in mission review.
Log PSR entries with both shadow-onset and shadow-exit quilt patches. The shadow-onset patch gives you a high-signal boundary where the quilt transitions from orbital DEM to acoustic prior; the shadow-exit patch is the ground truth you use to calibrate the quilt against post-PSR orbital imagery. Over a VIPER-scale campaign of four PSR excursions, this calibration tightens the acoustic-prior confidence model quickly, so later excursions push deeper with less margin.
For Artemis III architects scoping crewed PSR EVAs, use the acoustic quilt to pre-certify fallback paths. If the primary route becomes blocked mid-EVA, the crew needs an alternative that does not require real-time Earth consultation. EchoQuilt's patch-confidence field flags which fallback segments the planner trusts at EVA-start and which require additional data before they are committable. This turns PSR contingency planning from text-level ConOps into a data-driven decision, and it matches the formal verification methods already used for lunar shadowed regions.
When scouting a new PSR for the first time, capture an entry-direction acoustic baseline before any traverse commitment. The baseline is a 5-10 minute passive recording at the shadow boundary, which characterizes the local noise floor and the thermal-cycle acoustic signature of the regolith at the entry point. EchoQuilt uses this baseline to calibrate the patch-confidence field for the rest of the PSR; without it, the patch-confidence model assumes a generic noise environment and may either over-trust or under-trust patches inside the shadow. The baseline costs almost nothing in mission time and almost everything in subsequent traverse efficiency, so analog teams running PSR-analog tests at Mauna Kea or in cryogenic chambers should treat baseline capture as a non-negotiable step in their test protocol.
A cross-domain parallel comes from terrestrial bat hibernacula, where access-route planning uses the same acoustic-prior routing logic to plan biologist access to sensitive cave entries. Conservation biologists need to commit to access routes that minimize disturbance to roosting bats, and acoustic priors give them route confidence in the same way that planetary teams use them for PSR commitment. The methods have evolved in parallel and routinely benefit from cross-pollination of routing-confidence techniques.
Coordinate quilt updates with orbital DEM refresh cycles where possible. LROC narrow-angle camera images can be reprocessed against improved photometric models to produce updated DEMs at the PSR shadow boundary, and EchoQuilt's planner can re-fuse the acoustic quilt against the new DEM to update committed routes. The cadence of LROC updates is not aligned with rover sols, but the planner is built to accept asynchronous DEM updates and re-derive route confidence on the next supervisory window. This is one of the few places where a long-standing orbital asset like LROC directly improves a near-term traverse plan, and PSR programs that take advantage of it gain margin that programs treating the orbital data as static do not.
Ready to Route PSR Traverses With Acoustic Priors?
JPL and ESA mission planners preparing VIPER-class lunar rovers or Artemis III PSR excursions need a mapping tool that turns passive acoustic recordings into routable terrain priors. EchoQuilt's PSR integration is designed for that handoff, from scouting capture through route commit and post-excursion calibration. Each pilot ships with a battery-margin constraint module sized to the 60-90 minute usable PSR window, a parallel-pass scouting protocol that resolves local slope at 3-5 meter offsets, an entry-direction acoustic baseline configuration tuned for cryogenic chambers below 110 K, and an LROC narrow-angle camera DEM refresh integration that re-fuses orbital and acoustic priors on the next supervisory window. Pilot teams shape the pit-rim geometry extraction defaults and the metabolic-workload route-scoring weights that the 2027 Artemis III reference release will adopt.
Priority goes to JPL VIPER-heritage teams scoping follow-on lunar polar concepts, NASA Artemis III architect working groups planning crewed PSR EVAs within 2 km of landing sites, NIAC PIs targeting south pole water-ice extraction concepts, and ESA mission planners coordinating multi-agency PSR scouting cycles. Cryogenic-chamber analog test partners with confirmed 2026-27 field windows receive direct integration support from our analog field engineering team. Join the Waitlist for Planetary Analog Researchers to pilot the integration against your PSR-analog field tests and shape the quilt-to-planner interface before it locks.