PLR Embodiment Paper

Liquid handlers

📄Paper

Hamilton STARlet Full v0

Enabling high-throughput biology with flexible open-source automation

Chory EJ, Gretton DW, DeBenedictis EA, Esvelt KM

Molecular Systems Biology · 2021 · EMBO Press

DOI 10.15252/msb.20209942 ↗

Automated 480-culture turbidostats, M13 phage plaque assays, 100-condition metabolic profiling — all running asynchronously on a single Hamilton STARlet + BMG CLARIOstar. Demonstrates open-ended Python scheduling without pre-defined workflow graphs.

{'<'}/{'>'}Code

dgretton/many_asynchronous_turbidostats ↗ dgretton/pyhamilton_population_dynamics ↗ dgretton/roboplaque ↗

Three experiment-specific repos: (1) asynchronous turbidostat feedback loops, (2) simulated geographic population dynamics, (3) high-speed phage plaque assays. All use pyhamilton (MIT Media Lab) — PLR's Hamilton backend is built on the same VENUS/HSL bridge.

Uses pyhamilton, not PLR directly. PLR Hamilton v0 backend uses same approach.

📄Paper

Hamilton STARlet Full v0

Systematic molecular evolution enables robust biomolecule discovery

DeBenedictis EA, Chory EJ, Gretton DW, Wang B, Golas S, Esvelt KM

Nature Methods · 2022 · vol. 19 pp. 55–64

DOI 10.1038/s41592-021-01348-4 ↗

PRANCE: 96 parallel phage-assisted continuous evolution lagoons on one Hamilton STARlet. CLARIOstar luminescence feedback autonomously adjusts selection stringency every 30 min. Demonstrates evolution of T7 RNAP, tRNA synthetase, and tRNA simultaneously.

{'<'}/{'>'}Code

dgretton/std-96-pace ↗

Full PRANCE platform code: 96-well phage evolution scheduling, real-time luminescence feedback control, dilution/transfer logic. Self-contained experiment repo — not a library.

Also uses dgretton/roboplaque for plaque-based fitness assays.

📄Paper

Opentrons OT-2 Mostly v1

AssemblyTron: Flexible automation of DNA assembly with Opentrons OT-2 lab robots

Bryant JA Jr., Kellinger M, Longmire C, Miller R, Wright RC

Synthetic Biology · 2023 · Oxford University Press

DOI 10.1093/synbio/ysac032 ↗

Python package that reads j5 DNA assembly design files and executes them on the OT-2. Automates PCR with annealing gradient, Golden Gate assembly, and homology-dependent IVA cloning. Demonstrated on four-fragment chromoprotein reporter plasmid assemblies.

{'<'}/{'>'}Code

PlantSynBioLab/AssemblyTron ↗

Full OT-2 protocol package: j5 JSON → protocol script generation for PCR, Golden Gate, IVA. Reads standard j5 outputs (J5 Parameters, Assembly, Part Order CSV). End-to-end cloning automation.

Uses native Opentrons Python API; PLR's OT-2 backend wraps the same interface.

📄Paper

Opentrons OT-2 Mostly v1

An Open-Source Modular Framework for Automated Pipetting and Imaging Applications

Ouyang W, Bowman R, Wang H, Bumke KE, Collins JT, Spjuth O, Carreras-Puigvert J, Diederich B

bioRxiv · 2021

DOI 10.1101/2021.06.24.449732 ↗

Combines OT-2 liquid handling with a UC2 open-source modular microscope (Hi2, ~7 000 EUR). OT-2 moves samples; Raspberry Pi REST stack exposes pipetting and imaging via Jupyter notebooks. Automated sample prep + fluorescence imaging.

{'<'}/{'>'}Code

openUC2/UC2-Hi2 ↗

Full system repo: OT-2 REST control, UC2 microscope firmware, Jupyter workflow notebooks for combined pipetting + imaging. Modular design — OT-2 and microscope act as independent nodes.

Tecan Freedom EVO Basic v0 OPEN GAP

PLR-supported. Python wrappers exist (robotevo, pyTecan, PyEvo) but none are tied to a published peer-reviewed biology paper with open GitHub code. Best lead: robotevo (RNA virus detection, 2018 doctoral thesis — not a journal paper).

Plate readers

📄Paper

BMG CLARIOstar Full v1

Enabling high-throughput biology with flexible open-source automation

Chory EJ, Gretton DW, DeBenedictis EA, Esvelt KM

Molecular Systems Biology · 2021

DOI 10.15252/msb.20209942 ↗

CLARIOstar provides real-time absorbance/fluorescence feedback during turbidostat and plaque assay workflows. Python platereader wrapper reads results and feeds them into scheduling loop.

{'<'}/{'>'}Code

dgretton/platereader ↗

Lightweight Python interface to the CLARIOstar. Triggered from pyhamilton workflows — reads plates, returns raw absorbance/fluorescence data. 10-commit companion to pyhamilton.

📄Paper

BMG CLARIOstar Full v1

Systematic molecular evolution enables robust biomolecule discovery (PRANCE)

DeBenedictis EA, Chory EJ, Gretton DW, Wang B, Esvelt KM

Nature Methods · 2022

DOI 10.1038/s41592-021-01348-4 ↗

CLARIOstar is embedded inside the PRANCE enclosure. Luminescence reads from each of 96 phage-evolution lagoons drive closed-loop selection stringency control every 30 min.

{'<'}/{'>'}Code

dgretton/std-96-pace ↗

PRANCE platform code includes the CLARIOstar feedback loop — luminescence thresholds dynamically scale selection pressure. Real-time closed-loop molecular evolution.

📄Paper

Tecan Spark Full v1

Lustro: High-throughput optogenetic experiments enabled by automation and a yeast optogenetic toolkit

Harmer ZP, McClean MN

ACS Synthetic Biology · 2023

bioRxiv preprint ↗

Tecan Spark reads mScarlet-I fluorescence from 96-well yeast cultures under programmable LED illumination (optoPlate-96). High-throughput characterisation of optogenetic circuits in S. cerevisiae.

{'<'}/{'>'}Code

mccleanlab/Lustro ↗ mccleanlab/Optoplate-96 ↗

Lustro: Spark output analysis scripts + experiment orchestration. Optoplate-96: LED illumination firmware + scripts. Together they form a complete automated optogenetics platform.

📄Paper

Tecan Spark Full v1

Dynamic Multiplexed Control and Modeling of Optogenetic Systems Using Lustro

Harmer ZP, Thompson JC, Cole DL, Venturelli OS, Zavala VM, McClean MN

ACS Synthetic Biology · 2024

DOI 10.1021/acssynbio.3c00761 ↗

Extends Lustro to multiplex two split TF optogenetic systems (CRY2/CIB1 + eMagAF/eMagBF) with distinct light programs. Neural net + Bayesian optimisation identifies optimal light conditions from Spark fluorescence data.

{'<'}/{'>'}Code

mccleanlab/Optoplate-96 ↗ zavalab/ML / Optogenetics ↗

Two repos needed to reproduce: LED illumination control (Optoplate-96) and ML optimization scripts (Zavala lab). Tecan Spark reads are the experimental observations that train and validate the model.

📄Paper

Tecan Spark Full v1

Enhancing high-throughput optogenetics: Integration of LITOS with Lustro enables simultaneous light stimulation and shaking

Harmer ZP, Höhener TC, Landolt AE, Mitchell C, McClean M

microPublication Biology · 2024

DOI 10.17912/micropub.biology.001073 ↗

3D-printed adapter slots LITOS (Light-Induced Temperature-Oscillation System, Pertz lab) into the Lustro/Spark stack — allows LED stimulation while the plate shakes. Characterises mScarlet-I yeast reporter with simultaneous illumination and aeration.

{'<'}/{'>'}Code

pertzlab/LITOS ↗

LITOS illumination system firmware and scripts (University of Bern, Pertz lab). Integrates with Lustro's Spark-based measurement pipeline via the 3D-printed adapter.

BioTek Synergy H1 Full v1 OPEN GAP

PLR-supported with a full SynergyH1Backend. Paper found (Tang et al. 2023, immunomodulator HTS, DOI 10.1101/2023.06.26.546393) but no GitHub code published. SpectraMax i3x has a Python wrapper (GormleyLab) but no linked paper. Bio-Rad CFX96/384 has neither.

PCR / qPCR

📄Paper

AB QuantStudio 3/5 Not in PLR

Auto-qPCR: a python-based web app for automated and reproducible analysis of qPCR data

Maussion G, Thomas RA, Demirova I, Gu G, et al.

Scientific Reports · 2021

DOI 10.1038/s41598-021-99727-6 ↗

ΔCt / ΔΔCt / absolute quantification pipeline for QuantStudio 3/5 CSV exports. Applied to gene expression in iPSC lines, cortical and dopaminergic neuron differentiation, and cocaine-treatment datasets. Web app deployed on shinyapps.io.

Post-acquisition analysis tool, not real-time instrument control. Instrument explicitly named; real biology performed.

{'<'}/{'>'}Code

neuroeddu/Auto-qPCR ↗

Python + R Shiny web app: ingest QuantStudio CSV → normalisation → statistics → publication-ready plots. Used in the paper to reanalyse multiple independent qPCR datasets.

Bio-Rad CFX96/384 Not in PLR NO PAIR FOUND

No public Python API, no peer-reviewed paper with open GitHub code found. CFX Maestro software is closed. The biggest gap in the PCR category.

Imaging

📄Paper

MD ImageXpress Nano Pico in PLR

wrmXpress: A modular package for high-throughput image analysis of parasitic and free-living worms

Wheeler NJ, Gallo KJ, Rehborg EJG, Ryan KT, Chan JD, Zamanian M

PLOS Neglected Tropical Diseases · 2022

DOI 10.1371/journal.pntd.0010937 ↗

Python + CellProfiler pipeline for ImageXpress Nano plate imaging. Five analysis modes: optical flow (motility), Cellpose/YOLOv8 segmentation, CellProfiler morphology, Trackpy tracking, diagnostics. Applied to C. elegans, Brugia malayi, Schistosoma mansoni anthelmintic screens.

{'<'}/{'>'}Code

zamanianlab/wrmXpress ↗

Modular analysis package: wrapper.py parses HTD metadata from all ImageXpress models (Nano + Pico + Confocal). Five pluggable pipeline modes. Snakemake workflow integration. Actively maintained (Zamanian lab, UW-Madison).

Uses ImageXpress Nano; PLR driver targets the Pico — both use the same HTD file format.

📄Paper

MD ImageXpress Nano Pico in PLR

A graphical user interface for wrmXpress 2.0 streamlines helminth phenotypic screening

Caterer Z, Horejsi RV, Weber C, Mathisen B, et al., Zamanian M, Wheeler NJ

Int. J. Parasitol.: Drugs & Drug Resistance · 2025

DOI 10.1016/j.ijpddr.2025.100588 ↗

Dash-based GUI wrapping wrmXpress 2.0. New tracking pipeline (distance, velocity, tortuosity, directional change). Biology: S. mansoni miracidia praziquantel dose-response curves.

{'<'}/{'>'}Code

wheelerlab-uwec/wrmXpress-gui ↗ wheelerlab-uwec/GUI_ms ↗

Two repos: GUI app (Dash) and analysis scripts (R + raw data) for the paper's dose-response experiments. GUI_ms contains the praziquantel S. mansoni raw data and R analysis pipeline.

BioTek Cytation 5 Full v1 OPEN GAP

PLR driver (PR #277, Oct 2024) supports brightfield, fluorescence (DAPI/GFP/RFP/YFP/Cy5/Cy7) and autofocus. No peer-reviewed paper with GitHub Python automation code found as of 2026-10. Cytation 5 is widely used but papers rarely provide code.

Robot arms & dispensers

📄Paper

Brooks PreciseFlex 400 Full v1

GLAS: an open-source easily expandable Git-based scheduling architecture for integral lab automation

Cousty J-C, Cavagna T, Schmidt A, Mariano E, Villat K, de Nanteuil F, Miéville P

Digital Discovery · 2024 · RSC

DOI 10.1039/D4DD00253A ↗

Full automated chemistry lab at EPFL Swiss Cat+ facility. PreciseFlex SCARA arm moves plates/vials between instruments for overnight catalysis optimisation campaigns. Git-based scheduler with REST API, node modules per instrument, reaction screening + purification + analysis.

Domain: synthetic chemistry / catalysis. Arm model in paper is PreciseFlex 3400 (longer-reach SCARA variant; same family as PLR's 400 driver).

{'<'}/{'>'}Code

swisscatplus/glas ↗ swisscatplus/glas-web-client ↗

Orchestrator + instrument node modules + scheduling engine. PreciseFlex arm controlled via socket-level wrapper. Web client for scheduling and monitoring. Complete self-driving lab architecture.

Formulatrix Mantis WIP PR#987 OPEN GAP

PLR integration in-progress (PR #987, opened 2026-04). Mantis is widely used for NGS library prep but always via proprietary Formulatrix API — no Python + biology paper + GitHub code combination found.

UFACTORY xArm 6 Basic v1 TOO NEW

Driver merged September 2026 — no publications yet. xArm 6 is cost-effective and popular in university labs; biology-automation papers expected 2026–2027.

Peripherals & incubators

📄Paper

Liconic STX Full v1

Integrating Cell Painting and Thermal Proteome Profiling for mechanism-of-action profiling of bioactive compounds

Johansson C et al. (pharmbio, Uppsala)

bioRxiv · 2025

DOI 10.1101/2025.05.30.657006 ↗

AROS platform (UR10 arm, Liconic incubator, BioTek washer/dispenser, ImageXpress microscope). Automates Cell Painting + Thermal Proteome Profiling across 5,300 compounds in U2OS cells. Full self-driving biology lab pipeline.

Uses pharmbio/aros (predates PLR's Liconic backend added Mar 2026). Same hardware, different Python framework.

{'<'}/{'>'}Code

pharmbio/aros ↗ pharmbio/robotlab ↗ integrate-cp-tpp (analysis) ↗

aros: hardware orchestration for the full platform. robotlab: Liconic incubator Python driver. integrate-cp-tpp: paper analysis scripts (Cell Painting + TPP data integration). All repos are public.

📄Paper

BTX Gemini X2 Basic v1

Active learning guides automated discovery of DNA delivery via electroporation for non-model microbes

Crits-Christoph A et al. (Cultivarium)

bioRxiv · 2025

DOI 10.1101/2025.11.18.689155 ↗

Bayesian optimisation for electroporation parameter discovery across non-model microbes. Active learning loop selects voltage, pulse duration, media conditions. Custom automated electroporator built because BTX Gemini X2 couldn't cycle parameters programmatically.

BTX Gemini X2 is a comparison instrument in this paper, not the primary. Paper built a custom device to overcome its limitations. PLR's BTX driver (PR #1063) was not used.

{'<'}/{'>'}Code

cultivarium/electroporation-bayesian-optimization ↗

Full Bayesian optimisation pipeline for electroporation: surrogate model, acquisition function, automated experiment loop. Non-model organism panel (soil bacteria, environmental isolates). Self-driving biology design.

Thermo Cytomat 2/10 Full v1 OPEN GAP

PLR Full v1 support (2024). Cytomat is used in integrated platforms (e.g. Zhao Group, AD-SDL BIO_workcell at Argonne) but those use separate control frameworks. No biology paper with PLR's Cytomat backend + open GitHub code found yet. Papers expected from contributing labs (Adaptyv Bio, BioCam/Lance TUM).

AD-SDL BIO_workcell (Argonne) NOTABLE — NO PAPER YET

AD-SDL/BIO_workcell ↗ — Liconic STX88 + Azenta sealer/XPeel + Hudson SOLO + Hidex plate reader, controlled via ROS2/WEI framework. Code is public; no biology paper published yet. Closest thing to a full PLR-peripheral platform paper coming from a national lab.

Simulation & reference assets

What geometry and reference data exist for each PLR instrument? URDF (.urdf/.xacro), MJCF (.xml), meshes (STL/DAE/OBJ), labware JSON, and public manuals. Only public assets are listed. ⊘ = exists but requires vendor login or NDA.

Machine	PLR tier	URDF / xacro	MJCF (.xml)	Meshes STL/DAE/OBJ	Ref JSON / labware	Manual / spec
Robot arms — best-covered category
Brooks PreciseFlex 400	Full v1	RoboDrop/pf400_description ↗ .urdf + .xacro + .srdf (MoveIt) ⚠ 3rd-party — documented approximations	—	STL meshes (in pkg) ↗ Visual + collision geometry included	—	Brooks product page ↗
UFACTORY xArm 6	Basic v1	xArm-Developer/xarm_ros2 ↗ xarm_description/urdf/xarm6/xarm6.urdf.xacro Official UFACTORY · ROS2 + Gazebo + MoveIt2 xarm_ros (ROS1) ↗	— Community conversions circulating; not in mujoco_menagerie	STL + DAE meshes ↗ In xarm_description/meshes/ · visual + collision	xArm-Python-SDK ↗ Joint config, DH params, JSON API schemas	xArm 6 User Manual ↗ PDF — freely downloadable from UFACTORY
Universal Robots UR3/UR5/UR10	Full v1	UR_ROS2_Description ↗ urdf/ur.urdf.xacro · all models parameterised Official Universal Robots · calibration-aware ros-industrial/universal_robot (ROS1) ↗	mujoco_menagerie/ur5e ↗ mujoco_menagerie/ur10e ↗ Google DeepMind · validated against hardware Includes actuator params, contact geometry	STL + DAE (ROS2 pkg) ↗ OBJ meshes (menagerie) ↗ Production-quality, both visual + collision	RTDE interface guide ↗ Real-time data exchange schemas · JSON-compatible RTDE Python client ↗	UR Technical Specs ↗ DH params, payload curves, reach diagrams — all public
Liquid handlers — significant gap
Opentrons OT-2	Mostly v1	— No official URDF anywhere	★ opentrons-mujoco-viz ↗ Only public OT-2 kinematic model in any format Hand-authored ot2.xml from technical spec · Text2WetLab	— No official mesh files; community Thingiverse STLs for deck (cosmetic only)	shared-data/labware ↗ Official JSON defs · wells, volumes, offsets shared-data/deck ↗ OT-2 deck slot positions JSON	Opentrons support docs ↗ OT-2 User Manual + Python API reference
Opentrons Flex	Basic v1	—	—	—	shared-data/labware ↗ Same JSON labware library covers Flex decks shared-data/deck (Flex layout) ↗	Opentrons Flex docs ↗
Hamilton STAR / STARlet	Full v0	—	—	—	PLR Python resources ↗ Python-format well/rack definitions — no open JSON Native format: .rck / .bpl (VENUS, proprietary)	⊘ Hamilton VENUS manual — licensed, via Hamilton support
Tecan Freedom EVO	Basic v0	—	—	—	— EVOware worklist .gwl format is proprietary text	⊘ Tecan EVO User Guide — licensed / restricted
Plate readers, PCR & imaging — no sim assets exist
BMG CLARIOstar	Full v1	—	—	—	— Output: Neptune .nc / .xml (proprietary)	⊘ BMG LABTECH manuals — registration required
Tecan Spark	Full v1	—	—	—	—	⊘
Bio-Rad CFX Connect	Full v1	—	—	—	— Output: .pcrd (proprietary XML-based)	⊘
Applied Bio QuantStudio	Full v1	—	—	—	—	⊘
Molecular Devices ImageXpress	Full v1	—	—	—	— Output: .htd plate metadata (plate position/layout)	⊘
Peripherals & incubators — no sim assets exist
Liconic STX	Full v1	—	—	—	—	⊘
Thermo Cytomat 2/10	Full v1	—	—	—	—	⊘
Formulatrix Mantis	WIP	—	—	—	—	⊘

link = public asset amber = unofficial / community ★ = notable / only known ⊘ = exists but licensed / restricted — = no known asset

AICell Lab digital-lab workcell NOTABLE — ROS URDF WORKCELL

cccoolll/digital-lab ↗ — Closest public example of a full PLR-adjacent workcell with geometry: Dorna arm URDF + generic plate incubator URDF + Squid microscope model, all in ROS. Not PLR instruments by name, but same instrument class. Workcell URDF + STL assets are public. AccelerationConsortium/Matterix ↗ covers chemistry SDL in Isaac Sim / USD format.

Coverage gap: All PLR liquid handlers and peripherals have zero public URDF, MJCF, STL, or OBJ assets. Only the three robot arm families (PreciseFlex, xArm 6, UR3/5/10) have any simulation geometry. Universal Robots are uniquely well-served: UR5e and UR10e appear in Google DeepMind's mujoco_menagerie ↗ with production-quality MJCF + OBJ meshes, contact geometry, and validated actuator parameters. The opentrons-mujoco-viz OT-2 MJCF (ot2.xml) remains the only public kinematic model for any PLR liquid handler — hand-authored from the technical spec with no upstream geometry to convert from. Opentrons does publish rich labware definition JSON ↗ (well positions, volumes, offsets) which are the most complete open reference for any liquid handler deck layout.