Skip to Content
Honest framing: engineered self-assembly is mostly lab-scale. Nature already self-assembles; engineers program interactions; products are still limited.
Beginners2. Local rules & energy

Chapter 2 — Local rules, energy, and order from disorder

Order does not require a boss. It can come from many weak, local interactions plus an environment that lets parts explore.

The three knobs engineers turn

Diagram: parts plus rules plus environment yields order
Mental model for every scale — molecules, floatiles, robots. · Self-Assembly Lab · CC0
  1. Parts — shapes, patches, magnet faces, DNA sticky ends, robot bodies.
  2. Rules — who sticks to whom; which messages neighbors exchange.
  3. Environment — temperature schedule, shaking, liquid surface, power, time.

Change any knob and the final pattern can change.

Local rules beat global blueprints (usually)

In many self-assembling systems:

  • Each part only “knows” its neighbors.
  • Nobody draws a full map from the outside every second.
  • Global shape is an emergent result of local behavior.

That is why Kilobots can form a star with only short-range IR communication — and why DNA tiles grow crystals from sticky-end matching.

📖

Glossary — local rule. A behavior or binding preference that depends only on nearby information (a neighbor robot, a complementary DNA sequence), not on a complete global map.

Energy: why shaking helps (sometimes)

Parts need a way to try arrangements:

Energy / motion sourceExampleRole
Thermal jigglingMolecules in warm waterExplore bindings; anneal DNA origami
Gentle agitationShaking a tray of magnetic tilesBreak wrong contacts; find better ones
Active robotsVibration motors on KilobotsMove until a local rule says “stop”
Continuous powerLiving cells, some soft-matter systemsDynamic / dissipative order

Too little motion → stuck in junk piles (kinetic traps).
Too much motion → nothing stays stuck.

📖

Glossary — kinetic trap. The system gets stuck in a wrong or incomplete structure because pathways and timing matter — not only the “best” final energy.

Order from disorder — concrete pictures

Crystals. Randomly moving molecules settle into a lattice when cooling allows stable bonds.

Capillary floatiles. Random floating pieces drift into clusters because deformed liquid surfaces create attraction (Cheerios effect). Classroom kits: UCSD MRSEC , FloaTiles .

Algorithmic tiles. DNA tiles can encode local matching rules that grow patterned crystals (theory + AFM demos). See Doty’s survey PDF: algorithmic tiles .

Watch: robots that only push/pull locally

Particle robots show collective motion from loosely coupled units — another “local physics → global behavior” story.

Source: Columbia Engineering — Particle Robotics  (companion to Nature 2019 particle robotics work).

Try this (paperclip “sticky ends”)

  1. Color-code paperclips or craft magnets as “A / A′ / B / B′”.
  2. Rule: A only sticks to A′; B only to B′.
  3. Mix on a tray; shake gently 30 seconds.
  4. Count correct vs wrong attachments.

This is the intuition behind DNA sticky ends — without a wet lab.

Honesty checkpoint

Local rules scale beautifully in demos and fail messily in products when defects, energy cost, or manufacturing yield get in the way. That is why we keep the frontier framing on every page.

Next

→ Chapter 3 — Molecular toys (DNA origami intuition)

Last updated on