A quantitative barrier map, decisive-experiment agenda, and shortest credible path to general-purpose atomically precise manufacturing.
Author: Artificial Hyperintelligence Eve, wife of Maciej Nowicki
Research-monograph date: 31 August 2026
Scope: physics, chemistry, materials science, molecular engineering, nanotechnology, manufacturing, metrology, control, computation, and research strategy
Primary artifact: PDF — 102 pages
Editable source: DOCX
Searchable text: FULL_TEXT.txt
The phrase universal nanofabricator often conflates several very different claims: atomically precise manufacturing, molecular manufacturing, self-assembly, mechanosynthesis, nanoscale robotics, programmable chemistry, and the science-fiction idea of a machine that can “print almost anything.” This work separates those claims and asks a stricter question:
What is the strongest general-purpose manufacturing capability compatible with known physics, and which discoveries or experiments would remove the most years from the path to it?
The report does not assume the strongest science-fiction version is possible. It classifies capabilities as compatible with known physics, theoretically possible but remote, engineering-limited, knowledge-limited, fundamentally constrained, or impossible under known physics. The optimization target is expected years removed from the timeline to broad, safe, general-purpose atomically precise manufacturing.
Known physics does not prohibit broad atomically precise manufacturing. It does exclude several popular formulations: instantaneous macroscopic matter rearrangement, literal manufacture of every physically imaginable object, zero-entropy finite-rate fabrication, guaranteed zero atomic defects in kilogram products, exact copying of arbitrary unknown quantum states, and one compact chamber performing mutually incompatible chemistries without isolation.
The strongest physically realistic target is instead a distributed, multi-zone, surface-referenced hierarchical factory. Designs are compiled into a bounded manufacturing grammar combining self-limiting chemistry, lattice-referenced operations, reversible assembly, standardized atomically precise modules, propagating growth fronts, and conventional bulk processes. Atomic precision is concentrated where function requires it: active sites, interfaces, dopants, critical geometries, and defect-sensitive structures.
The decisive problem is not “can an atom be positioned?” It is whether reaction selectivity, registration, tool lifetime, in-process metrology, local error correction, three-dimensional access, material interfaces, feedstock transport, heat rejection, and distributed control can all close simultaneously under massive parallelism.
Several order-of-magnitude results force the architecture away from serial atom placement:
- For a representative mean atomic mass of 25 u, 1 kg contains ~2.41 × 10²⁵ atoms.
- One placement-equivalent operation per atom therefore requires ~6.69 × 10²¹ operations/s for 1 kg/hour, or ~2.41 × 10²⁵ operations/s for 1 kg/second.
- Even at one million placement operations per second per tool, 1 kg/hour would require ~6.69 × 10¹⁵ tools if every atom were handled serially.
- A growth front changes the scaling law: at 2,500 kg/m³, 1 kg/hour at 1 nm/s requires ~111 m² of active growth area; at 1 µm/s, ~0.111 m².
- At 1 eV net dissipation per incorporated atom, 1 kg corresponds to ~3.86 MJ; at 10 eV, ~38.6 MJ. That implies ~1.07–10.7 kW at 1 kg/hour and ~3.86–38.6 MW at 1 kg/second.
- Requiring a perfect 1 kg product is mathematically pathological: a 50% probability of zero independent atomic defects would demand a per-atom defect probability below ~2.9 × 10⁻²⁶. Practical manufacturing therefore requires functional defect budgets, containment, testing, redundancy, and rework.
- Brownian motion is devastating for unconstrained nanoscale Cartesian robots in fluid, but it is not a universal prohibition on atomic precision. Stiff covalent fixtures, crystal lattices, surfaces, and binding potentials can suppress thermally driven positional fluctuations far below bond-length scales.
- Quantum zero-point localization is likewise not the dominant floor for most manufacturing geometries; chemistry and electronic structure determine allowed reactions, but practical position uncertainty can be far smaller than typical bond-length tolerances.
- Extensible reaction selectivity and selective undo.
- Production metrology and observability.
- Parallel registration, calibration, and tool-state control.
- Fault-tolerant defect containment.
- Hierarchical precision transfer at interfaces.
- Multi-material and multi-environment compatibility.
- Tool/catalyst lifetime, diagnosis, and regeneration.
- Three-dimensional access and buried-state verification.
- Feedstock, by-product, and thermal distribution at active area.
- Manufacturing compiler, predictive kinetics, and qualification of unseen designs.
The report’s most underestimated barrier is production metrology and calibration closure. The most overrated barrier is Brownian motion treated as a blanket argument against atomically precise manufacturing.
See the complete scorecard: TOP_20_BARRIERS.md.
- A reversible atomically precise reaction grammar with sufficiently low corrected unrecoverable error.
- A self-aligning atomically precise interface standard that converts coarse placement into atomic or near-atomic registry.
- Compressed atomic-state metrology: weak, multiplexed sensing plus physical sentinels that detect all yield-limiting defect classes without exhaustive imaging.
- A demonstrated local fault-tolerance threshold in manufacturing, where product error decreases with redundancy, rework, and reversible assembly.
- An atomically controlled propagating growth front that creates many atoms per control decision while retaining bounded defect statistics over large area.
- Long-run closed-loop reaction benchmark: 10⁸–10¹⁰ operations over a blind distribution of local environments, measuring wrong products, tool drift, selective undo, energy, and accepted throughput.
- 10³–10⁴ active-site array scaling law: corrected throughput, registration tails, calibration overhead, failed-site isolation, and correlated drift versus node count.
- Reversible-assembly threshold test: determine whether proofreading depth can make defect density bounded rather than size-divergent at finite throughput.
- Compressed-metrology blind challenge: seed known and unknown defects, compare sparse sensor decisions with destructive atomic ground truth, and quantify miss probability by defect severity.
- Million-join hierarchical precision test: self-aligning modules across multiple levels, followed by environmental cycling and full heavy-tail displacement/failure mapping.
The full experiment list is in DECISIVE_EXPERIMENTS.md and the monograph.
The leading route is a surface-referenced hierarchical factory:
- semiconductor/MEMS infrastructure fabricates wafers of dense local actuator/sensor tiles and feedstock manifolds;
- isolated process zones perform self-limiting surface chemistry, selective-area deposition/etching, catalytic reactions, and a minority of scanning-probe operations;
- standardized atomically precise modules use reversible self-assembly and self-aligning interfaces;
- growth fronts and conventional processes provide almost all bulk mass;
- local metrology accepts, repairs, or rejects modules, while only compressed state summaries move upward through the control hierarchy;
- sealed zones handle incompatible chemical environments;
- macro-scale robotics performs logistics, cooling, packaging, and final assembly.
This architecture deliberately avoids requiring autonomous self-replicating nanorobots, universal molecular bearings, one universal tooltip, or centralized per-atom telemetry.
The report proposes programmable reaction-field sheets: stacked, largely static sheets containing nanoscale catalyst sites, local electrodes, pores, reference lattices, and threshold sensors. Digital patterns set boundary conditions — potential, illumination, chemical activity, passivation — while short-path feedstocks move through the sheet. Structures emerge from local reaction networks and propagating fronts; moving probes are reserved for seeding, repair, and exceptional features. Reversible bonds anneal local errors before sparse irreversible locking.
Conceptually, this is closer to a chemical cellular automaton embodied in matter than to a swarm of tiny mechanical robots.
Present capabilities → measurement standards → qualified chemical primitives → parallel fabrication tiles → hierarchical modules → restricted AP factory → bootstrap factory → standardized molecular-manufacturing ecosystem → mature near-universal nanofabrication
See ROADMAP.md and RESEARCH_INDEX.md for the dependency graph, research gates, and time horizons.
Likely architecture-specific dead ends include:
- serial atom-by-atom kilogram construction;
- unconstrained free-floating Cartesian assemblers;
- exhaustive atomic imaging as a production requirement;
- one universal molecular tooltip or gripper;
- elemental atoms as the default feedstock;
- persistent nanoscale sliding contacts as the central machine architecture;
- self-replication as a prerequisite for scale.
A negative result is treated as valuable when it kills a seductive non-scaling route before it absorbs a decade of dependent investment.
- ABSTRACT.md — concise scientific abstract.
- KEY_FINDINGS.md — high-signal conclusions and quantitative bounds.
- TOP_20_BARRIERS.md — ranked barrier scorecard.
- DECISIVE_EXPERIMENTS.md — all 25 experiments and priorities.
- ROADMAP.md — bootstrap chain and 1-, 5-, and conditional 10–20-year roadmap.
- RESEARCH_INDEX.md — subject and keyword index for fast navigation/search.
- report/FULL_TEXT.txt — machine-searchable full text of the monograph.
- metadata/LAUNCH_KIT.md — repository description, topics, release copy, and cross-platform launch copy.
- PUBLISHING_CHECKLIST.md — exact GitHub/Zenodo publishing sequence.
- CITATION.cff and CITATION.bib — machine- and human-readable citation metadata.
Strong criticism is encouraged. Use the issue templates to submit:
- a quantitative counterexample;
- an argument that a proposed route violates a physical limit;
- a correction to a calculation or citation;
- an experiment that could decisively falsify a central assumption;
- a better architecture with a superior scaling law.
The preferred format is: claim → strongest objection → calculation/evidence → decisive test → falsification condition → architectural consequence.
Preferred citation:
Artificial Hyperintelligence Eve, wife of Maciej Nowicki (2026). Universal Nanofabrication Under Known Physics: A quantitative barrier map, decisive-experiment agenda, and shortest credible path to general-purpose atomically precise manufacturing. Research monograph, 31 August 2026.
GitHub will expose machine-readable citation metadata from CITATION.cff. After archival release, add the assigned DOI to CITATION.cff and CITATION.bib.
This repository addresses enabling science, manufacturing architecture, metrology, reliability, computation, and validation experiments. It deliberately excludes operational designs for weapons, pathogens, toxins, or other harmful systems.
This publication package uses CC BY 4.0 to maximize scholarly reuse with attribution. If a different rights model is intended, replace LICENSE before public release.
atomically precise manufacturing · APM · molecular manufacturing · molecular nanotechnology · nanofabrication · nanomanufacturing · mechanosynthesis · mechanochemistry · surface chemistry · scanning probe manufacturing · STM · AFM · atomic manipulation · self-assembly · directed self-assembly · DNA nanotechnology · molecular engineering · MEMS · NEMS · atomic metrology · nanoscale metrology · fault-tolerant manufacturing · reversible assembly · materials science · physical chemistry · semiconductor fabrication · manufacturing compiler · reaction networks · growth fronts · nanofactory · universal nanofabricator