Source Review: Walther et al. 2024 — Autonomous Construction with In-Situ Boulders

Summary

Claim 1: "Construction throughput approximately 174-193 m² per operational day"

Verdict: Novel supporting Why: Provides a concrete throughput benchmark for autonomous lunar construction using one excavator. Useful for sizing the mobility/construction component of our calc and for projecting M7 (manufacturing complement operational) timing under different fleet sizes. For our 31.5 t mobility fleet, 20 rovers × 175 m²/day ≈ 3,500 m²/day throughput aggregate — well above the rate needed to build a few quarter-ring blast shields and landing pads in a single year.

Claim 2: "1,440-1,520 operational hours to build a quarter-ring blast shield segment"

Verdict: Novel supporting Why: Direct evidence that even with a single excavator, lunar surface construction is achievable in tens-of-days timescales rather than years. Supports IE-regime time-compression for the construction-phase milestones (M3 habitat deployment, M4 FSP installation, M5 ISRU plant pilot) which all involve some surface construction.

Claim 3: "An excavator capable of operating in the lunar environment needs to be developed"

Verdict: Consistent Why: Confirms the hardware-development is not yet at deployment TRL — consistent with our calc's BAU baseline assumption that we are at TRL 4-6 for these components.

Claim 4: "Autonomous path planning using A* algorithm, greedy targeting, payload management, and terrain slope constraint enforcement"

Verdict: Not relevant (technical detail) Why: Algorithm-stack details. Useful as evidence that the autonomy problem is well-bounded and solvable, but no direct numerical claims for q5.

Cross-reference

• The in-situ-boulders-only construction path is the lowest-mass option for surface infrastructure — uses zero Earth-imported regolith or processed materials.
• Most useful as evidence that the construction-throughput problem is bounded and tractable, not as a direct cost benchmark.
• Pairs with Metzger-Autry 2022/2023 to bracket "what does lunar construction cost as a function of fleet design choice."
• Codex anti-hallucination check: all quoted text appears verbatim in the extract.md.

Abstract

Verbatim: "Significant infrastructure is required to establish a long-term presence of humans on the lunar surface. In-situ resource utilization (ISRU) is a fundamental approach to ensure the viability of such construction. Here, we investigate the feasibility of constructing blast shields as one example of lunar infrastructure using unprocessed lunar boulders and an autonomous robotic excavator." The study models a 10 m³ payload-capacity excavator deploying in-situ boulders to construct quarter-ring blast shield segments. Construction throughput is ~174-193 m² per operational day; one quarter-ring (~314 m perimeter) requires 1,440-1,520 operational hours (~75-80 days at 20 hrs/day operation). The path-planning combines A* with greedy targeting under terrain-slope constraints. The paper does not provide a capital-cost estimate, an excavator-design TRL, or a system-level mass-to-surface number; it is a feasibility study at the planning-algorithm and throughput-arithmetic level.

Key claims

• in-situ-boulders-construction: "An autonomous robotic excavator can use unprocessed lunar boulders to construct quarter-ring blast shield segments"
• throughput-174-193-m2-day: "Construction throughput approximately 174-193 m² per operational day"
• quarter-ring-1440-1520-hrs: "1,440-1,520 operational hours to build a quarter-ring blast shield segment"
• 10m3-payload-excavator: "10 m³ payload capacity for boulder collection"
• excavator-needs-development: "An excavator capable of operating in the lunar environment needs to be developed" — i.e. the hardware is not at deployment TRL today
• automation-stack: "Autonomous path planning using A* algorithm, greedy targeting, payload management, and terrain slope constraint enforcement"

Reviewer notes

A purely demonstrational study — establishes that the autonomy-stack is solvable, but explicitly defers the hardware-development question. Load-bearing for q5 at a qualitative level: shows that in-situ-boulders-only construction (no regolith processing, no sintering) is achievable with one well-designed excavator, which is a much lower-mass / lower-capex path than processed-regolith methods (compare metzger-autry-2022-landing-pads' microwave sintering). Notably absent: cost in dollars, mass landed, TRL assessment, deployment timeline. Useful as evidence that the construction-throughput problem is bounded and tractable, not as a cost benchmark. Cross-leaf relevance: the lower-capex construction path matters for q5 buildup scenarios where Earth-launch mass dominates.