Skip to main content

SpikerBot (Pre-Release)

SpikerBot combines a brain design studio with a palm-sized neurorobot, letting learners assemble spiking neural circuits into brain models that see, hear, move, blink, and beep.

Built for classrooms. SpikerBot translates complex brain science into an affordable, hands-on robot that lets students model neural circuits, see neural activity drive behavior, and build skills that underpin modern neurotechnology, nervous system medicine, and neuromorphic AI.

Get the App​

Download the SpikerBot App — available for Windows, macOS, iOS, Android, and Chrome OS.

When the app opens, you’ll see a brain-shaped canvas. Drag black (excitatory) or white (inhibitory) neurons onto the canvas, wire them up, and press Play to run the network.

Design Mode (Paused)​

App Design Mode Screenshot

Live Brain Mode​

App Live Mode Screenshot

Hardware at a Glance​

SpikerBot is powered by an ESP32-S3 microcontroller with Wi-Fi and Bluetooth, running on four AA batteries. Its sensor suite includes:

  • 320 × 240 forward-facing RGB camera
  • Microphone (mono, 8-bit PCM)
  • Infrared distance sensor (~1 m range)

Output devices include:

  • Two differential drive motors (left and right wheels)
  • Four independently addressable RGB LEDs
  • PWM tone speaker with PCM audio playback

The robot communicates wirelessly with the app via 2.4 GHz Wi-Fi for real-time control.

SpikerBot Robot

(AI-generated hardware preview)

Sensor / EffectorSpecification
RGB Camera320 × 240 px, forward-facing
Distance SensorInfrared (IR), approx. 1 meter range
MicrophoneMono, 8-bit PCM stream
MotorsDifferential drive (left/right wheels)
RGB LEDsFour individually programmable LEDs
SpeakerPWM tone + PCM audio output
MCUESP32-S3 dual-core with Wi-Fi + BLE
Power4 × AA batteries

The ESP32 handles all sensing and actuation, while the neural simulation runs on your paired device and sends commands to the robot in real time.

Getting Started​

  1. Power On: Flip the switch located on the underside of the robot.
  2. Connect: On your computer or tablet, connect to the robot’s Wi-Fi network (SSID printed on the robot).
  3. Wait: The robot’s LEDs will turn green when ready.
  4. Launch the App: Open the SpikerBot app, load or build a brain, and press Play to start controlling the robot.

Robot App

Inside the App​

The SpikerBot App provides an interactive workspace shaped like a brain where you design embodied spiking neural networks.

  • Drag & Drop: Add excitatory (black) or inhibitory (white) neurons onto the canvas.
  • Connect: Draw synaptic connections by dragging from neuron outputs (axon handles) to neuron inputs or effectors.
  • Tune: Click any neuron to adjust properties such as firing type (Quiet, Burst, Custom) and polarity (Excitatory/Inhibitory).
  • Play: Run the network and watch your robot respond in real time.
  • Save: Brain designs can be saved or loaded in a portable JSON .brain format for sharing or further editing.

Neurons fire discrete spikes simulated by the biologically inspired Izhikevich model. Synapse weights (1-100) set connection strength: weights over ~25 ensure reliable activation; weights near 90+ create persistent feedback loops (similar to short-term memory).

Neurons 1

Neural Circuit Building Blocks​

Common neural circuit motifs you’ll use include:

  • Reflex Arc: Direct sensor-to-effector connection producing simple stimulus-response behavior.
  • Crossed Circuit: Connects left sensor to right motor and vice versa, enabling turning behaviors.
  • Synaptic Integration: Multiple weak inputs converge on a neuron that fires only when inputs coincide.
  • Recurrent Excitation: Feedback loops maintain neuron activity beyond the initial stimulus.
  • Mutual Inhibition: Competing neurons inhibit each other, enabling decision-making by allowing only one behavior at a time.

Recurrent Excitation
Mutual Inhibition

How to Design Brains​

  1. Define: Specify the desired robot behavior in observable terms.
  2. Decompose: Break the behavior into smaller sub-tasks.
  3. Draft: Build neural circuits (motifs) corresponding to each sub-task.
  4. Inhibit: Use inhibitory neurons to prevent conflicting behaviors.
  5. Test and Refine: Observe robot behavior, identify issues, and iteratively improve your network.

Brain Library​

Starter Pack​

Kickstart your SpikerBot exploration with six pre‑configured example brains:

  1. Download the Starter Pack (all-brains.zip).
  2. Unzip the contents into your local Documents/SpikerBot/text directory.
  3. Load them in the app via File → Load, then choose any .brain file to run instantly.

Brain: Tracker

Tracker — Retinotopic Smooth Pursuit​

  • Neurons: 17 (all excitatory)
  • Function: Keeps a moving green object centered by mapping nine horizontal “retinal” zones to differential wheel speeds, producing smooth, continuous tracking rather than zig-zag turns.
  • Download: Tracker.brain

Brain: Two-Mood Pet

Two-Mood Pet — Bistable State Model​

  • Neurons: 30 (25 excitatory, 5 inhibitory)
  • Function: Simulates two behavioral “moods” (Curious vs. Sleepy) using mutually inhibitory recurrent loops, illustrating rapid state switching akin to biological sleep–wake mechanisms.
  • Download: Two-Mood Pet.brain

Troubleshooting​

SymptomSuggested Fix
Robot silent or immobileCheck power and battery status.
Wi-Fi won’t connectToggle Play in app, confirm SSID, wait for LEDs to turn green.
No movementEnsure wheels are attached properly and unobstructed.
Color mis-detectionImprove lighting conditions and check camera view.

Next Steps​

  • Experiment with synapse weights to feel the emergence of reliable and memory-like behaviors.
  • Combine neural circuits to build custom brains, then share your creations with the community!