Week 3 HW: LabAutomation

1. Reference paper: HYDRA – hydrogels by robotic automation

Citation

Torchia, E. et al. Fabrication of cell culture hydrogels by robotic liquid handling automation for high-throughput drug testing. Communications Engineering, 4, 222 (2025).

What they did

This paper introduces HYDRA (HYDrogels by Robotic liquid-handling Automation), a method to fabricate thin, planar hydrogel films directly inside standard 96- and 384-well plates using liquid-handling robots.

Normally, when you cast hydrogels into small wells, capillary forces at the sidewalls create a curved meniscus, which:

  • makes hydrogel thickness non-uniform across the well
  • disturbs cell seeding density and imaging focus
  • reduces the reliability of high-throughput drug screening

HYDRA solves this by:

  1. Robotically dispensing a sub-contact volume of hydrogel precursor (fish gelatin + microbial transglutaminase) into each well, carefully avoiding contact with the sidewalls.
  2. Immediately re-aspirating most of that volume with precisely controlled height and flow rate.
  3. Using contact angle hysteresis so that a thin, meniscus-free layer (about 10–50 μm) remains at the bottom.

The authors show that these hydrogels support drug dose–response assays on engineered epithelial cells and allow long-term imaging on soft, biomimetic substrates. They also demonstrate that HYDRA can be implemented on an open-source Opentrons OT-2 robot, effectively turning a liquid-handling platform into a simple, programmable soft-materials fabrication tool.

They combine this with plate-scale quality control and show that the hydrogels support:

  • Drug dose–response assays with engineered epithelial cells
  • Long-term holographic and fluorescence microscopy on soft, biomimetic substrates

How the automation is implemented (and why it’s relevant)

  • They explicitly use an Opentrons OT-2 as an open-source, low-cost platform to implement HYDRA. The pipeline is built with Opentrons Protocol Designer and custom Python to control dispense/aspirate heights, volumes, and speeds. :contentReference[oaicite:4]{index=4}
  • The OT-2 mixes gelatin and transglutaminase stocks, casts the precursor into plates, and re-aspirates to leave a controlled film.
  • Conceptually, this turns a “liquid-handling robot” into a materials fabrication tool: it is designing not only concentrations but also geometry (flat films with controlled thickness) and mechanical properties (tunable stiffness ~1.5–6 kPa). :contentReference[oaicite:5]{index=5}

For my interests (digital fabrication, soft metamaterials, auxetics), this is very close to “2.5D soft material printing”:

  • Recipe space: gelatin / crosslinker concentrations
  • Process parameters: volumes, pipette height, flow rate
  • Output behavior: thickness, flatness, stiffness and cell response

This is exactly the kind of workflow I want to adapt: using Opentrons not just as a biology helper, but as a programmable fabrication device for soft, structured materials.

Planned automation project: Opentrons-printed auxetic hydrogel tiles

Working title

Opentrons-printed auxetic hydrogel tiles for programmable mechanics

Core idea

Use the Opentrons OT-2 as a “dot-matrix printer” for soft materials: it will deposit small droplets of hydrogel precursor with different formulations onto a thin flexible substrate, forming a 2D auxetic (negative Poisson’s ratio) pattern.

By controlling which beams/tiles are soft vs stiff (or swell more vs less), the overall structure exhibits programmable shape change or auxetic behavior when stretched or stimulated.

This combines:

  • HYDRA’s idea of robotic hydrogel fabrication
  • My background in digital fabrication and mechanical metamaterials
  • A design space of structure × material formulation × process

Biological / material system (high-level)

I will use a crosslinkable hydrogel system compatible with HTGAA and the Opentrons:

  • Option A: fish gelatin + microbial transglutaminase (following HYDRA)
  • Option B: a photo-crosslinkable gelatin (e.g., GelMA) if lab infrastructure favors photogels

Each “unit” in the auxetic pattern is defined by two main parameters:

  1. Base stiffness – controlled by gelatin concentration (e.g., 5%, 10%, 20%)
  2. Crosslink density – controlled by enzyme concentration or light exposure time

These combinations create:

  • Soft segments that deform easily
  • Stiff segments that act as constraints

Arranged in an auxetic geometry (e.g., re-entrant honeycomb, rotating squares), the global behavior becomes a tunable mechanical metamaterial.

Cells are optional at this stage; the primary goal is to demonstrate programmable mechanical behavior. Cells could later be seeded to test how different stiffness regions affect attachment and morphology.

What will be automated

  1. Automated formulation library

    • The OT-2 prepares a small library of hydrogel precursor formulations in a 96-well plate:
      • “Soft”: 5% gelatin + 0.5% crosslinker
      • “Medium”: 10% gelatin + 1% crosslinker
      • “Stiff”: 20% gelatin + 2% crosslinker
    • Using standard pipetting commands, the robot mixes stock solutions to create these recipes in defined wells.
    • This step establishes a recipe space that can be expanded later (different polymer, different additives, etc.).

  2. Geometric pattern → robot coordinates

    • I design an auxetic pattern in Rhino/Grasshopper or Python (e.g., 6 × 6 re-entrant lattice).
    • Each structural element (“beam”) is discretized into a small number of print points.
    • Each point is annotated with a recipe ID (“soft”, “medium”, “stiff”).
    • I then map this point set into the OT-2 deck coordinate system by:
      • 3D-printing a simple holder that clamps a thin transparent membrane (PDMS or plastic) on the deck
      • Calibrating its four corners relative to the robot origin
      • Converting design coordinates (x, y) into deck positions via a linear transform + offsets

    The result: a table of points like
    (x, y, recipe_id) that the robot can iterate through.

  3. Printing and curing

    • The robot:
      • Aspirates small volumes (e.g., 2–5 µL) of each formulation from the 96-well “recipe plate”
      • Moves to specified (x, y) positions over the substrate
      • Deposits droplets in the order dictated by the auxetic pattern
    • After printing:
      • The membrane is transferred to an incubator (37 °C) or UV station for crosslinking
      • Once cured, the membrane can be mounted in a simple tensile frame and mechanically tested (even by hand with phone video) to observe auxetic deformation.

This is a direct analog of HYDRA (controlling meniscus and layer thickness), but applied to patterned, multi-formulation structures instead of uniform coatings.

Experimental workflow (high-level)

  1. Preparation (manual)

    • Prepare gelatin and crosslinker stock solutions.
    • Design and 3D-print a membrane holder compatible with a specific Opentrons deck slot.
    • Attach a thin transparent membrane on top and calibrate four reference points on the OT-2.

  2. Deck layout

    • Slot 1: Eppendorf rack with stock solutions (gelatin, crosslinker, buffer).
    • Slot 2: 96-well “recipe plate” where the robot generates soft/medium/stiff formulations.
    • Slot 3: custom membrane holder (“printing bed”).
    • Slot 4–5: tip racks (P20).
    • Slot 6: waste reservoir.

  3. Opentrons protocol (concept)

    • Step 1 – Formulation generation

      Robot mixes gelatin + crosslinker into the recipe plate:

      • Use P300/P20 to combine stocks in wells A1–A3 to produce “soft”, “medium”, “stiff”.
      • Optional: generate more variants across the plate to explore stiffness/swell space.

    • Step 2 – Auxetic pattern printing

      • Load auxetic point list (x, y, recipe) from a CSV or embedded list.
      • For each point:
        • Pick up a tip
        • Aspirate 2 µL from the well corresponding to recipe
        • Move to the transformed (x, y, z) above the membrane
        • Dispense the droplet
        • Drop the tip
      • Repeat until all beams in the pattern are printed.

    • Step 3 – Curing and testing

      • Move the printed membrane to an incubator / UV lamp for crosslinking.
      • After curing, mount on a simple frame, apply tension, and record deformation.
      • Optional: overlay grid or markers to track local strain.

    My Final Project – Possible Directions

1. Opentrons-printed auxetic hydrogel tiles

One-liner
Use an Opentrons robot as a “2.5D printer” to deposit hydrogel droplets in an auxetic pattern, and study how composition plus geometry shape the mechanical behavior.

What I would automate

  • Opentrons mixes a small library of hydrogel formulations in a 96-well plate (for example, soft / medium / stiff).
  • It then “prints” droplets onto a thin flexible film in a precomputed auxetic pattern (re-entrant squares, rotating squares, etc.).
  • Each beam or node of the pattern can use a different formulation, so the auxetic response is not only geometric but also rheology-informed.

Why it is interesting

  • Treats the OT-2 as a soft-material fabrication machine instead of just a pipetting tool.
  • Links soft-matter rheology, mechanical metamaterials, and lab automation: geometry, composition, and process are all programmable.
  • Can start as a purely mechanical experiment and later add biological layers if there is time.

2. High-throughput hydrogel rheology map in a plate

One-liner
Use lab automation to build a small materials map of hydrogel rheology in a 96-well plate, linking formulation to mechanical properties as a design tool for later printing.

What I would automate

  • Opentrons prepares a combinatorial grid of hydrogel recipes across the plate
    (for example, rows = polymer concentration, columns = crosslinker concentration).
  • For each well, I run a simple, automatable mechanical proxy
    (for example, indentation depth under a fixed weight, or image-based deformation).
  • Collect all measurements into a composition–property map that I can use to choose formulations for the auxetic printing in project 1.

Why it is interesting

  • Turns subjective “this gel feels soft or stiff” into structured data driven by automation.
  • Shows a clear automation loop: the robot explores a soft-material design space, not just prepares biological assays.
  • Directly supports and informs the first project idea.

3. Droplet-printed bio-logic patterns (cell-free logic gates)

One-liner
Use automation to assemble and place small cell-free reactions that behave like simple logic gates, so that a 2D fluorescent pattern encodes a truth table in space.

What I would automate

  • Opentrons prepares cell-free reactions with different DNA constructs that approximate AND / OR / NOT behavior (or simpler ON / OFF variants).
  • Each well corresponds to an input combination (00, 01, 10, 11), arranged in the plate as a visual truth table, or optionally printed as droplets onto a flat substrate.
  • After incubation, the pattern of fluorescence across wells or positions becomes a spatial representation of the logic.

Why it is interesting

  • Uses automation to build and arrange many small reactions that would be tedious by hand.
  • Connects synthetic biology, simple computation, and digital fabrication: logic is expressed both in biochemical reactions and in spatial layout.
  • Offers a complementary direction where the “printed pattern” carries information and function, not just mechanical behavior.