Skip to content

Hrushimanju/DATOM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 

Repository files navigation

DATOM — Modular Robotic Programmable Matter

MSc Dissertation Project | University of Liverpool | 2024
Supervised by Dr. Paolo Paoletti | School of Engineering


What is this?

DATOM (Deformable Autonomous Transformable Modular) is a self-reconfiguring robotic programmable matter system. Individual modules connect, communicate, and reconfigure autonomously to form adaptive 3D structures.

This project was supervised by Dr. Paolo Paoletti at the University of Liverpool. The primary goal was to develop between one and five functional DATOM modules capable of inter-module communication, magnetic latching and detaching, and forming demonstrable shapes or patterns as proof of concept.

My contribution was the design and physical build of a single DATOM module — the necessary foundation before multi-module communication could be achieved.


What was built

  • Full dodecahedral ABS shell designed in Creo Parametric through multiple design iterations
  • Scissor-lift actuation mechanism selected and modelled after ruling out linear actuators due to volume constraints
  • Arduino Nano-based control system with IR inter-module communication
  • Dual magnetic latching system — permanent neodymium magnets for passive connection, electromagnets for active switching
  • Power management system: 9V input, dual voltage regulation (5V/6V), MOSFET-controlled electromagnet switching. Total system draw of 9.55W determined from component datasheet specifications
  • IR signal transmission verified via oscilloscope waveform analysis in controlled lab conditions

Key engineering decisions

Volume constraint — the central problem

Linear actuators require a retracted length equal to half their extended length. Inside a dodecahedral shell with fixed internal dimensions, this makes them geometrically infeasible.

Decision: Switched to scissor-lift mechanism. Provides vertical actuation within a footprint that fits the shell geometry. Motor mounts alongside the mechanism rather than in series with it.

Component selection for miniaturisation pathway

Every component was selected with size as a primary constraint:

  • Arduino Nano over Uno — smaller footprint
  • Tower Pro micro servo — fits within connector housing
  • BS170G MOSFET — minimal board space for electromagnet switching
  • IR communication over Bluetooth/WiFi — no external module required

Material selection

ABS was selected over PLA based on material research and datasheet comparison. ABS offers higher impact resistance and better dimensional stability under mechanical stress — relevant properties for module-to-module connection forces and magnetic latching loads. This conclusion was research-based rather than derived from physical comparative testing.


Honest results

IR communication between modules was demonstrated in controlled lab conditions. Oscilloscope analysis confirmed signal transmission. Noise interference under real operating conditions was documented as an unresolved variable — the root cause was identified as requiring hardware-level shielding rather than software filtering alone. This remains an open engineering problem for the next iteration.

An informal component layout test — placing all selected components inside a half-scale shell — confirmed spatial feasibility at reduced scale. However this was not formally documented and actuator load requirements under operational forces were not calculated. It is possible that a correctly sized actuator introduces additional volume or structural constraints that invalidate this informal test. Formal load calculations and validated component fit remain outstanding work.


Technical insights

  • Designing for volume constraints forces better component decisions than designing for performance alone
  • IR communication noise in embedded systems at this scale requires hardware-level shielding — software filtering alone is insufficient because the signal-to-noise ratio degrades below recoverable thresholds in confined, multi-component enclosures
  • Datasheet specifications assume isolated operating conditions. Real system draw, thermal behaviour, and signal integrity diverge once components share a confined volume — this gap is where most embedded integration failures occur
  • Informal spatial feasibility checks are necessary but not sufficient. The component fit passed visual inspection at half-scale, but without actuator load calculations, any volume margin could be consumed by a correctly rated motor

Project status

Phase 1 — Complete: Single DATOM module designed, built, and tested. Dodecahedral shell, scissor-lift actuation, Arduino Nano control, IR communication verified, dual magnetic latching system integrated. IR noise and actuator load calculations identified as open problems.

Phase 2 — Requires fabrication access: PCB design to replace breadboard circuit (volume reduction + noise shielding), formal load calculations for scissor-lift under operational module weight, reduced-scale shell reprint with validated component fit, multi-module IR communication in shielded environment, autonomous latching sequence with IR handshake protocol. Target: 3–5 communicating modules forming demonstrable shapes.


Current parallel work

While DATOM miniaturisation requires fabrication resources currently unavailable, I am building a low-cost predictive maintenance vibration monitor that addresses two open DATOM problems directly:

  1. Latching verification — vibration frequency analysis can confirm whether magnetic latching between modules was successful or partial. Failed or misaligned connections produce distinct frequency signatures that can trigger re-attempt sequences.

  2. Actuator wear detection — the scissor-lift mechanism will degrade under repeated cycling. Vibration monitoring of the actuator provides early warning of mechanical wear before failure, extending operational lifespan of each module.

Both problems require the same core skills: embedded sensor integration, real-time signal processing, and noise handling in constrained environments — the exact challenges identified in the DATOM IR communication work.


Tech stack

Area Tools
Mechanical CAD Creo Parametric (primary design and iteration), SolidWorks (STL export, print-readiness validation via Ultimaker Cura)
Prototyping FDM 3D printing (ABS), hand assembly
Microcontroller Arduino Nano (ATmega328P)
Communication IR (SFH4545 emitter, TSOP382 receiver)
Power management 7805/7806 regulators, BS170G MOSFET
Verification Oscilloscope waveform analysis
Documentation Technical report, BOM, circuit diagrams

References

Full dissertation report available on request.
Supervisor: Dr. Paolo Paoletti, University of Liverpool School of Engineering.

Key reference: Piranda, B. and Bourgeois, J. (2022) "Datom: A Deformable Modular Robot for Building Self-reconfigurable Programmable Matter"


This project is part of an ongoing personal roadmap toward functional programmable matter systems. Questions and collaboration welcome.

About

Self-reconfiguring modular robotic system — MSc Dissertation, University of Liverpool 2024

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors