Builds

Hardware & Lab Engineering

From Drosophila sleep-tracking arrays to hydroponic grows and unconventional computing prototypes — hands-on engineering across domains.

Ethoscope Array
6
Neuroscience

Ethoscope Array

High-throughput Drosophila behavior and sleep tracking system. Ethoscopes use machine vision and real-time tracking for long-duration experiments with minimal manual intervention.

Raspberry Pi-based tracking units
Infrared illumination for dark-cycle recording
Machine vision tracking algorithms
Multi-unit parallel experiment support
PiVR System
4
Neuroscience

PiVR System

Raspberry Pi Virtual Reality system for closed-loop optogenetic and thermogenetic experiments on freely moving insects. Enables real-time stimulus delivery based on animal position.

Real-time animal tracking at 30+ fps
Closed-loop LED/thermal stimulus delivery
Custom 3D-printed enclosures
Python control software
Lab Engineering

Peristaltic Pump System

Custom-built peristaltic pump for precise fluid delivery in biological experiments. Designed for controlled drug administration in ethomics setups.

Stepper motor-driven peristalsis
Programmable flow rate control
Arduino/RPi microcontroller interface
Multi-channel capability
Shallow Water Culture Grow
10
Agriculture

Shallow Water Culture Grow

Hydroponic shallow water culture (SWC) system for cannabis vegetation. Optimized for root oxygenation, nutrient uptake, and rapid vegetative growth.

Aerated nutrient reservoir
Net pot suspension system
pH and EC monitoring
LED grow lighting integration
Agriculture

Seedling Propagation Station

Controlled-environment seedling propagation system for starting clones and seedlings under optimized light, humidity, and temperature conditions before transplanting to vegetative growth systems.

Humidity dome propagation trays
T5 fluorescent and LED lighting
Temperature-controlled heating mats
Misting and clone nutrient protocols
Microtubule Cymatic Reservoir Computer
3
Unconventional Computing

Microtubule Cymatic Reservoir Computer

A GPU-native, neuroplastic Language Reservoir Model (LRM) built to simulate biological cytoskeletal dynamics. It features a 6,500-dimensional, 4-layer hierarchical architecture (Φ-Fast, Φ-Mid, Φ-Slow, Σ-Record) that utilizes Hebbian and Anti-Hebbian learning alongside Intrinsic Plasticity. The system uses Jaeger Conceptors to govern intentional state spaces, allowing it to translate chaotic real-time telemetry into coherent semantic responses.

6,500-dimensional 4-layer architecture
Hebbian + Anti-Hebbian learning with Oja's Rule
Intrinsic Plasticity (gain/bias adaptation)
Conceptor-gated semantic generation