Make the encounter tractable.
Use fixtures, selective chemistry, interfaces, or surfaces to reduce the relevant degrees of freedom. An incoming component need not hold a precise orientation while freely diffusing.
01 / A scientific field worth building · 2026
Molecular
Workcell
Engineering.
From molecular components we can build to molecular operations we can trust.
A proposed scientific discipline for reliable, verifiable, composable work at molecular scales. Its horizon is programmable molecular manufacturing and, ultimately, matter whose structure and function can be engineered with extraordinary precision.
An independent field-building proposal by Tristan Roberge-Mentec. Illustration is conceptual, not a reported experiment.
Conceptual system / MWE–001Architecture study
Molecular-scale operation
01 / ConstrainUse molecular geometry and local chemistry to increase the probability of the intended interaction.
Not a smaller robot arm.Work with thermal physics, not against it
Not a single device.A composable class of molecular operations
Not a promise of APM.A way to test the path toward it
02 / The central thesis
We can build parts.
Can we build processes?
We can fold DNA into intricate structures, design molecular machines, drive nanoscale motion, and synthesize atomically defined products in specialized settings. Yet these achievements do not automatically add up to a system that can perform reliable sequences of useful molecular operations.
Molecular Workcell Engineering asks how to create local environments where positioning, selective chemical action, measurement, and correction can be engineered together, then connected into longer workflows.
The unit of progress is not a beautiful nanoscale device. It is a verified operation whose performance survives contact with the next operation.
The current fragmentation
Components without a common engineering language.
Structural DNA nanotechnology, molecular motors, synthetic chemistry, sensing, and nanofabrication advance on separate metrics. Each has impressive demonstrations. Few experiments test their compatibility as an end-to-end process.
The proposed discipline
Functions that can be measured, compared, and composed.
Develop shared benchmarks for controlled encounters, selective action, state readout, error accounting, and multi-step reliability. Let multiple architectures compete on the same tasks, instead of presuming a single path will win.
03 / A distinct unit of engineering
The workcell
is the missing unit.
Conventional molecular research often reports what a structure is or a machine can do. Molecular Workcell Engineering asks what a complete operation does reproducibly, under specified conditions, with its failure modes accounted for.
That is a proposed common engineering discipline, not a claim that one particular scaffold or robot must win. The physical environment may do most of the work.
Convert state into useful work.
Carry out a bounded attachment, chemical transformation, displacement, or materialization step. Test whether the action produces the intended product rather than merely movement.
Measure, reject, or proceed.
Establish how the readout maps to actual outcome. Account for misattachment, incomplete transformation, and false verification before connecting the next operation.
Not a shrunken industrial robot
The tool, scaffold, energy supply, and measurement may be distributed across molecules, an instrument, and an externally controlled environment.
Not a single preferred architecture
DNA, designed proteins, synthetic motors, surfaces, scanning probes, and solution chemistry can be compared on the same verified-operation task.
04 / Research agenda
Seven fronts.
One discipline.
The field is deliberately larger than DNA origami, artificial molecular machines, or surface chemistry alone. It needs a common experimental language across physical regimes while preserving competing architectures.
The founding question: what must be measured, controlled, and reproduced before a molecular action becomes an engineering primitive?
01 / FIX↗
Programmable fixtures
Addressable geometries, confinement, local orientation, substrate interfaces, and selective attachment that increase useful encounters without requiring impossible mechanical authority.
02 / INTERFACE↗
Hybrid molecular interfaces
DNA, proteins, peptides, enzymes, and nonbiological surfaces used as interoperable structural and functional elements. Compatibility, cross-talk, and fouling are first-class variables.
03 / DRIVE↗
Energy and actuation
Chemical, electrochemical, optical, and magnetic ways to bias molecular events. The question is not whether motion occurs, but whether it contributes useful, selective work.
04 / READ↗
State sensing and metrology
Single-molecule observability, optical and spin-sensitive reporters, and calibrated inference. A signal must be causally tied to the operation it claims to verify.
05 / RECOVER↗
Error-aware operations
Misconnection, incomplete reactions, off-target events, and repair or discard pathways. Reliability must be measured through failure distributions, not best-case demonstrations.
06 / MODEL↗
Predictive design systems
Stochastic dynamics, molecular simulation, experimental design, automated analysis, and uncertainty-aware digital models that are tested against measurements rather than treated as proof.
07 / EXTEND↗
Beyond biological substrates
Explore transitions from biomolecular workcells to inorganic materials, surface-bound synthesis, and atomically defined structures. A long-term field must confront substrate transfer and industrial scale rather than assume biological scaffolds are universally suitable.
Deliberate constraint: no preferred architecture wins by definition. Solution-phase chemistry, self-assembly, and established lithography are legitimate baselines. Molecular workcells must demonstrate a capability or performance advantage worth their additional complexity.
05 / The long horizon
From programmed
molecules to
programmed matter.
Atomic precision is an ambitious target, but it is not a sensible requirement for every component. The field should choose fabrication methods by the function they deliver, the fidelity they require, and their full cost of verification.
A credible bridge to atomically precise manufacturing may be hybrid: bulk or template-directed assembly creates most of an object, while chemically selective, high-precision methods create or repair the small fraction of features that determine its performance.
Programmable biomolecular assembly
Addressable DNA and protein structures, recognition, conformational logic, and selective catalysis.
Bio-templated materials
Transfer encoded geometry or function into inorganic material through selected deposition and growth processes.
Hybrid precision fabrication
Combine molecular templates with surfaces, lithography, field control, and site-specific processing.
Atomically selective operations
Explore positional chemistry where exact local composition or bonding is indispensable, subject to throughput and metrology constraints.
The economic test: increase precision only when the resulting functional advantage exceeds the penalties of yield loss, instrumentation, precursor preparation, contamination control, and verification. General-purpose molecular manufacturing remains unproven, and these regimes are not stages every technology must traverse.
06 / What would count as proof?
One operation.
Fully accounted for.
The first common challenge should be deceptively simple: bring a component to an intended site, direct a selective transformation or attachment, and independently verify what happened. Then run enough trials to characterize the errors.
A second test asks whether the verified output can become the input to another operation without sacrificing reliability.
A proposed Workcell Benchmark
Measures, not promises01 / Outcome
Correct-operation yieldFraction of attempts producing the intended outcome, with uncertainty and independent confirmation.
02 / Selectivity
Off-target eventsWrong binding, unintended chemistry, interference, and failure to reject near-matches.
03 / Measurement
Verification fidelityFalse-positive and false-negative rates; correspondence between reporter signal and ground truth.
04 / Dynamics
Time and throughputOperation cycle time, usable event rates, and the dependence on conditions and scale.
05 / Resources
Energy and costEnergy, reagents, instrumentation, and effective cost per verified successful operation.
06 / Scale
Compositional reliabilityHow performance changes after operations are connected and independently replicated.
Why this matters: with independent uncorrected steps of 99.9% success, the probability of a flawless 1,000-step sequence is
0.999^1000 ≈ 36.8%. Correlated errors may be worse. Verification, recovery, and alternative process designs are not optional extras.GATE / 01
Selective event
Define an operation and demonstrate it under measured conditions.
GATE / 02
Independent replication
Separate teams reproduce the operation and its error profile.
GATE / 03
Verified state
Show a readout with known sensitivity, specificity, and measurement cost.
GATE / 04
Composable sequence
Connect distinct steps and quantify the cumulative failure rate.
GATE / 05
Useful task
Beat a conventional alternative in a well-defined application.
07 / Reasons this could fail
Make the hard
problems public.
The field should be organized around problems capable of killing its own thesis, rather than protected by an ever-expanding technology roadmap. A candidate workcell must outperform the relevant simpler alternative.
Three questions are especially decisive. The technical companion lays out eleven testable bottlenecks and proposed acceptance criteria.
Did the intended attachment actually happen?
Correctly docked, misattached, and unattached states can be indistinguishable in a conventional bulk measurement.
Evidence required: independently calibrated readout with misattachment and false-positive rates reported per operation.
Can verification scale with fabrication?
A process that requires bespoke microscopy for every object may never be economically useful even when its chemistry works.
Evidence required: lower-cost or parallel readout validated against an independent measurement with defined detection limits.
Do the engineering rules actually transfer?
If every machine needs new bespoke rules, there may be no coherent field or shared design system, only a loose collection of devices.
Evidence required: a predeclared rule improves first-pass functional performance in a second, genuinely different machine class.
08 / The field-building horizon
Build the field,
not just a lab.
A nascent field needs more than inspiring devices. It needs a shared definition of progress, a place to publish negative results, new collaborations, and the beginnings of a durable funding and training structure.
These are proposed milestones, not existing programs or guaranteed outcomes.
YEAR 01 / CONVENE
Define the first trials
Publish a short technical manifesto and a versioned benchmark; award small exploratory grants; host an open methods workshop across molecular machinery, nanometrology, simulation, and chemistry.
YEAR 02 / REPLICATE
Turn demonstrations into datasets
Run multi-lab round-robin tests; release baseline measurements and negative results; form cross-disciplinary teams around operations instead of material platforms.
YEAR 03 / COMPOSE
Make interoperability a research target
Fund experiments that connect separately developed fixtures, actuation, chemistry, and state readout; publish performance failures as well as successes.
YEAR 04 / INSTITUTIONALIZE
Create a recognizable research home
Establish a recurring Molecular Workcell Trials meeting, interdisciplinary training material, collaborative funding lines, and a dedicated journal collection or proceedings track.
YEAR 05 / TEST UTILITY
Show which operations matter
Demonstrate independently reproduced multi-step processes, evaluate manufacturability, and identify the applications where workcells outperform simpler methods. Retire architectures that do not.
09 / Scientific basis
Ambition should
survive evidence.
Workcell Engineering is a proposed integration discipline. The individual technical pillars are real; their reliable and economical composition remains an open research problem.
Demonstrated in published work
Parts of the foundation exist.
Programmable DNA scaffolds; synthetic molecular machines carrying out selective chemistry; electrically and magnetically actuated DNA devices; RNA-fueled motion; modeling tools for hybrid nucleic-acid nanostructures. These results are important but are not equivalent to a general molecular factory.
Scientifically unresolved
The integration gap is serious.
How to combine useful force and chemical selectivity with reliable placement, direct verification, recovery from defects, high throughput, and tolerable costs. Molecular state reporters, including spin-sensitive DNA systems, are promising avenues, not proven general workcell sensors.
Long-horizon hypothesis
Programmable matter remains the destination.
Atomically precise systems, adaptive artificial substrates, and ultimately self-maintaining structures could become accessible if verified molecular operations can be repeated and composed. No present experimental result establishes that such general capability is achievable or economical.
Selected primary literature / Technical anchors
[01]
Pumm et al. — A DNA origami rotary ratchet motor. Nature, 2022. ↗
[02]Weiß et al. — Magnetic DNA origami nanorotors. Advanced Materials, 2026. ↗
[03]Sequence-selective decapeptide synthesis using two artificial molecular machines. JACS, 2021. ↗
[04]Bohlin et al. — Design and simulation of DNA, RNA and hybrid protein–nucleic acid nanostructures with oxView. Nature Protocols, 2022. ↗
[05]A tunable autonomous RNA-fueled micro-engine. Nature Communications, 2026. ↗
[06]Meng et al. — Optically addressable and programmable spins in DNA. bioRxiv preprint, 2026. ↗
These references illustrate independently developed scientific capabilities. They do not imply that their authors endorse, participate in, or are affiliated with this proposal. The 2026 spin-DNA result is a preprint, not peer-reviewed published work.
10 / Who belongs in this field?
An uncommon coalition around a common problem.
The missing links are often between disciplines rather than inside them. The field should create an incentive for people who would otherwise never share experimental standards to work on the same measurable task.
DNA nanotechnologyProtein designSynthetic molecular machinesSurface chemistryNanometrologyQuantum sensingStatistical physicsControl engineeringAutonomous laboratoriesAtomic-scale fabrication
11 / Field proposer
Tristan
Roberge-Mentec
Independent research program architect
Frontier R&D · molecular systems · scientific AI
I work on translating ambitious, uncertain technical ideas into research programs with measurable hypotheses, scientific decision gates, and realistic routes to execution. My interests span molecular machinery, programmable biomolecular systems, atomically precise manufacturing, advanced materials, and AI-enabled experimental science.
My background includes technology-transfer work at CERN, commercialization and research-funding work at the University of Waterloo, and interdisciplinary R&D program integration at Zeden. My molecular workcell research is a conceptual and systems-engineering program, not a claim that I personally operate a validated molecular manufacturing laboratory.
The goal of this initiative is to convene the experimentalists, theorists, instrument builders, and research funders who could rigorously test whether molecular workcells deserve to become a field.
17Invention disclosures generated through an R&D program
5Provisional patent filings from that program
$400K+Non-dilutive funding secured or led
Professional figures describe prior R&D program work; they are not results from Molecular Workcell Engineering.
A field worth testing
The goal is not to claim a molecular factory. It is to discover how one could become possible.
Interested in the physics, chemistry, instrumentation, or research strategy? The field will be stronger if its assumptions are tested by people with different approaches.
Start a conversation