HiRISE 1 m
MOLA 463 m
A True Story · Apollo 14 · February 1971
In February 1971, two Apollo 14 astronauts climbed toward the rim of Cone Crater. The terrain was rugged, their orbital maps were inadequate, and every ridge looked the same. They turned back. Orbital images later showed they had stopped just 30 meters from the rim. Hope is not a navigation strategy. With no GPS, up to a 22-minute one-way signal delay, and global dust storms, the first explorers on Mars cannot afford a 30-meter mistake.
The same resolution ladder now compares , and before an reaches its . The shows what coarse terrain hides.
Our Approach
Every value carries its source. Missing data stays missing. Planning limits are labeled as planning limits, never as safety limits.
Three Scientific & Architectural Innovations
Curiosity end-of-drive IMU tilt points prove the resolution ladder matters. V4 stayed within 10 m of JPL's human-planned drive 45% of the time versus 41% on MOLA-only, weaving around >15° scarps.
Try it in the consoleScrub the modeled dust season and watch the planning status change.
The map names what it does not know: seam bias, archive lag, missing coverage and sub-meter rock limits.
Try it in the consoleMulti-Objective Pareto Proof · Gale to Escalante (2,018 ms)
In 2,018 ms, deterministic multi-objective Dijkstra generates six candidates over NASA terrain and cataloged USGS, IAU and evidence. Five form the non-dominated frontier; the shortest path is mathematically dominated by the cubic .
| Candidate Route | Pareto Status | Distance | Total Climb | Est. EVA | Mean / Max Slope | Trade-Off Takeaway | |
|---|---|---|---|---|---|---|---|
| 1. Low distance | Dominated in this set | 1,464.72 km | 31,332 m | 630.11 h | 3.54° / 31.23° | 94.3 | The shortest-path trap: minimizing horizontal distance forces 31.3 km of vertical climbing and the highest risk proxy. |
| 2. Low EVA | Pareto / non-dominated | 1,465.20 km (+480 m) | 22,597 m | 623.74 h | 3.05° / 32.90° | 93.1 | Walking just 480 meters farther (+0.03%) saves 8,735 meters of vertical climb and 6.37 hours of EVA time. |
| 3. Low terrain-risk proxy | Pareto / non-dominated | 1,510.34 km (+45.6 km) | 21,416 m | 639.83 h | 2.07° / 28.07° | 79.5 | Trading +45.6 km (+3.1% distance) cuts 9,916 meters of climb and lowers the cubic slope and roughness risk proxy from 94.3 to 79.5. |
| 4. Low terrain burden | Pareto / non-dominated | 1,620.55 km | 20,099 m | 681.72 h | 1.58° / 33.81° | 82.1 | Lowest total terrain burden (12.4 versus 23.9) and lowest cumulative climb, 11,233 m less than the shortest path. |
| 5. Science first | Pareto / non-dominated | 1,471.76 km | 22,815 m | 626.73 h | 2.22° / 32.90° | 85.9 | Weighted 55% Science, 20% Terrain, 15% EVA and 10% Ops to maximize proximity to cataloged USGS, IAU and THEMIS evidence. |
| 6. Balanced | Pareto / non-dominated | 1,474.38 km | 22,652 m | 627.58 h | 2.18° / 32.90° | 85.9 | Equal weights across all four normalized objectives. Score: 1.031. |
Panorama: NASA/JPL-Caltech/MSSS · Curiosity Mastcam, sol 4,982 (Aug. 11, 2026), yardang layer on Mount Sharp · Source. Move the cursor to look around and wipe the dust.
Inspect All 6 Pareto Candidates Live in ConsoleWhy This Matters Right Now (2026)
covers less than 2% of Mars. The x scale-bias model and 7.3 m seam detector flag coarse-map blind spots.
Analog crews and Artemis planners can rehearse multi-objective trade-offs by bridging coarse LOLA grids to 1 m LRO NAC strips.
Unifies 76 NASA PDS, USGS , ASU , MSL and NASA Ames MGCM sources in a browser console.
The Proof
Curiosity tilted on ground times steeper than the global map shows. On 3 m terrain the is : the map matches what the rover felt.
: from NASA MMGIS end-of-drive waypoints. Validation run . Rovers drive, they do not walk, and rocks under 1 to 2 m are below the terrain models. Read the full method
Every status reads Within Planning Limits, Review Required or Exceeds Planning Limit. No route or ground receives a safety verdict.
Where NASA data does not reach, the console says n/a or Unavailable. A gap is shown as a gap, never filled with a guess.
A deterministic engine owns every route and number. answers through and labels each answer as AI interpretation.
NASA Space Apps Challenge 2026
the best map possible
Terrain resolution ladder; synchronized 2D, USGS geology and 3D maps.
HiRISE, CTX, MOLA; USGS SIM 3292; IAU Gazetteer
details of their routes and destinations
MARSWALK planner, Route Design, multi-stop EVAs, comparison, EVA brief and exports.
MOLA, HiRISE, CTX; MMGIS rover traverses
updates on current conditions
One active sol drives modeled dust, archived RAD dose, rover weather, THEMIS history and live close approaches, each labeled.
NASA Ames Mars GCM, MSL RAD, REMS, MEDA, THEMIS, JPL CAD
the data necessary to make informed and scientifically meaningful decisions
Planning statuses with plain reasons, sensitivity, site science and evidence classes. NeuroNexus shows the data behind each decision and never labels ground safe.
All of the above, with a source tag on every value
NASA MOLA, MRO HiRISE and CTX, Mars Odyssey THEMIS, MSL RAD and REMS, Mars 2020 MEDA, NASA Ames Mars GCM, NASA MMGIS rover traverses, NASA/JPL Close-Approach Data, NASA Image and Video Library, USGS Astrogeology and the IAU Gazetteer, USGS SIM 3292, ESA Mars Express, Esri ArcGIS Mars services. See how each one is used
Team
