Autonomous excavator constructs a six-metre-high dry-stone wall
An autonomous excavator developed at ETH Zurich can scan irregular boulders and stack them into a six‑metre‑high dry‑stone wall, reviving an ancient construction technique without mortar or concrete. Commenters praise the robotics and planning algorithms needed to handle random rock geometry, but question the stability, aesthetics, and real‑world competitiveness of such walls versus gabions, shotcrete, or conventional reinforced concrete. The exchange broadens into how automation may reshape heavy construction, from safety and labor impacts to the thermodynamic and environmental advantages of building with unprocessed stone.
Overall Reaction to the Autonomous Excavator
- Many see this as an impressive robotics milestone: handling irregular, heavy stones is far harder than standardized bricklaying.
- Others criticize the demo wall as ugly, “sketchy,” or less tight-fitting than a skilled human’s work.
- Several note that traditional dry-stone work is a rare, expensive craft; automating it could be economically impactful if quality and safety match human standards.
Wall Quality, Stability, and Drainage
- Concern that the demonstrated wall is a single-thickness stack with little apparent batter or hearting; some doubt it could handle lateral loads.
- Others emphasize that dry-stone retaining walls should move and “breathe” slightly; rigidity is a bug, not a feature.
- Shotcrete is suggested for reinforcement, but multiple comments argue this:
- Can block drainage and effectively create a dam.
- Defeats the low-CO₂, no-concrete goal.
- May crack due to differential expansion vs rock.
- Alternatives like gabions and geogrid/tiebacks are brought up as more robust and easier to engineer.
Algorithms, Sensing, and Precision
- System reportedly scans ~20–30 rocks at a time, builds a digital twin, and optimizes stone placement for stability.
- Some feel the algorithm looks “greedy” and could do better with larger stone pools and more global planning.
- Discussion over whether centimeter accuracy is enough; consensus in-thread is that if the algorithm accounts for tolerance, cm-level is acceptable.
Engineering, Materials, and Comparisons to Concrete
- Debate over whether irregular stone structures can ever match concrete’s predictable, specifiable strength.
- Some suggest augmenting autonomous dry-stone with:
- Non-destructive rock-strength estimation (acoustics, vision).
- Tension rods to handle tensile loads.
- Others argue that for many infrastructures, automation plus stone could reduce energy use vs “pulverize rock → add cement → re-glue it.”
Labor, Robotics, and Future Use Cases
- Potential seen in reducing human labor for heavy, repetitive, high-skill excavation and wall-building.
- Idea of using the system as an “assistant” (e.g., AR guidance, construction logging) rather than full replacement.
- Broader speculation about autonomous machinery reshaping construction, landscaping, and infrastructure, with some caution around safety and real-world robustness.