The Journey So Far

This desktop robot began as an idea to combine sensors, movement, and personality into a palm-sized machine — one that could navigate its environment and display emotions through simple expressive cues.

Core Platform: ESP32 Brain

The robot’s main controller is an ESP32-WROOM-32, chosen for its dual-core performance, built-in Wi-Fi and Bluetooth, and flexible I/O support. Its processing power allows simultaneous control of motors, sensors, and display outputs — all while maintaining network connectivity for future OTA updates and data logging.

The ESP32 handles:

PWM motor control for precise movement

I²C communication with all sensors and OLED displays

Battery monitoring and power management

Serial debugging and configuration

Drive System: Two Wheels + Caster

Mobility is achieved through a two-wheel differential drive system powered by N20 500RPM micro gear motors, providing a balance between torque and speed.
Steering is accomplished by varying the speed of each motor — no servo steering is required.

To stabilize the chassis, a single rear caster (ball-style) supports the frame, keeping the bot balanced while allowing smooth pivot turns.

Power System

The bot runs on a 3.7 V 2000 mAh Li-ion battery, offering solid runtime for its small form factor.
A TP4056 USB-C charging and protection module manages charging and discharge safety. The power rail splits to feed:

ESP32 logic (3.3 V regulated)

Motor driver board (5 V or VIN depending on configuration)

Sensor and display I²C bus (3.3 V)

The goal is to later integrate automatic charging via magnetic pogo pins or a simple docking plate.

Motor Driver

A compact dual-channel PWM motor driver connects to the ESP32. Each motor is controlled via one PWM and one direction pin. This allows proportional speed control and smooth turning maneuvers.

Sensors: Environmental Awareness

The robot’s “eyes” are four VL53L0X Time-of-Flight sensors, placed strategically around the chassis:

Front-left and front-right for collision avoidance

Left and right sides for wall following and spatial mapping

Rear or angled sensor (optional) for backing awareness

Each sensor uses unique I²C addresses, set dynamically at startup to share the same bus.
An accelerometer (connected via I²C) provides tilt detection and motion feedback — letting the bot understand its orientation or detect if it’s been lifted or knocked.

Displays: Expression and Feedback

Two 0.96″ OLED displays (128×64, I²C) are mounted at the front.

Display 1: System data (battery level, IP, sensor readings, etc.)

Display 2: Animated “eyes” — giving the bot a bit of personality

Both displays share the same bus (address 0x3C), but a multiplexer or secondary I²C address can be used to differentiate them.

Layout and Wiring

Everything fits around a custom 3D-printed chassis, designed to roughly match the footprint of an Arduino Uno — compact but with layered sections for electronics and sensors.
The current wiring includes:

ESP32 GPIOs for both motor channels

Shared I²C for OLEDs, ToF sensors, and accelerometer

Power distribution from the TP4056 board

Motor driver VIN from battery rail

Battery sense pin for voltage monitoring

Software and Control Logic

Initial sketches handle:

Basic motor testing (PWM forward/reverse)

I²C detection of sensors and displays

Data visualization on OLED

Serial feedback for debugging

The next software milestones include:

Sensor fusion for obstacle detection

Motion logic with avoidance behavior

Display animations for idle/active states

OTA update integration

Wi-Fi dashboard for sensor readouts

Chassis Design

The chassis design is being modeled in Fusion 360, with considerations for:

Two motor mounts with alignment brackets

Rear caster housing

Front plate for ToF sensors and OLEDs

Detachable top shell for battery and board access

Clean internal wire routing

charging dock contact point, modular sensor mounts, or detachable head unit for easier prototyping.