The Sesame Robot Project

Greetings, from your new best friend.

Sesame is an accessible Open-Source robotics project based on the ESP32 microcontroller system, with an emphasis on expression and movement. This project is designed for makers and engineers of all skill levels! Sesame offers a dynamic platform designed to start working with walking robots. To build a sesame robot, you will need basic soldering skills, $50-60 in hardware components, access to a 3D printer, and a basic understanding of Arduino IDE.

This repository contains the CAD design files, STL files, build and wiring guides, and the base/expanded firmware for the ESP32-based controller. There is also some included debugging firmware that may be helpful in getting your Sesame up and running.

Features

• Quadruped Design: Uses 8 servo motors (2 per leg) to achieve roughly 8 total degrees of freedom.
• Emotive Display: Features a 128x64 OLED screen acting as a reactive face that syncs with movement.
• Fully Printable: Designed entirely for 3D printing in PLA with minimal supports.
• Network Connectivity: Connect to your WiFi network for remote control and API access.
• JSON API: RESTful API for programmatic control from Python, JavaScript, and more.
• Conversational Faces: Expressive emotion library with talk variants for voice assistant projects.
• Sesame Studio: New animation composer software to easily create custom movements.
• Sesame Companion App: Python application for voice control and advanced interactions.
• Serial CLI: Control the robot and trigger animations via a Serial Command Line Interface or the web UI.
• Pre-programmed Emotes: Includes animations for Walking, Waving, Dancing, Pointing, Resting, and more.

Watch the launch video on YouTube

Getting Started

Follow these steps to build your own Sesame Robot:

1. Gather Parts

Check the Bill of Materials (BOM) for a complete list of required electronics and hardware. * Microcontroller: Lolin S2 Mini (recommended for DIY builds), Sesame Distro Board V3 (Current, pre-flashed, supports Bambu Lab battery), V2 (legacy, USB-only), or ESP32-DevKitC-32E with Distro Board V1 (legacy) * Actuators: 8x 180 Degree MG90 Servos * Power: 5V 3A source (USB-C PD for S2 Mini and V2 Distro Board, or battery + buck converter; see BOM for the Bambu Lab 14500 7.4V 800mAh Li-ion Battery option)

2. Print Parts

Download the STLs and follow the Printing Guide. * Designed for PLA * Minimal supports required

3. Build & Wire

Follow the Build Guide and Wiring Guide to assemble the frame and connect the electronics.

4. Flash Firmware

Upload the code from the Firmware Directory. * Requires Arduino IDE * Configure WiFi AP settings

5. Create Animations

Use Sesame Studio to visually design poses and sequences for your robot.

Software & Firmware

Sesame Studio

Sesame Studio is a desktop application for designing animations, sequencing poses, and creating OLED expressions. It is maintained in its own dedicated repository at dorianborian/sesame-studio. It allows you to: * Visually pose the robot in a 3D viewport using the actual printed STL geometry. * Sequence frames into complex animations with custom transitions and delays. * Design 128x64 OLED face bitmaps or directly open, edit, and save Arduino face-bitmaps.h files. * Generate C++ code for servo angles and bitmaps ready to paste into firmware.

> Go to Sesame Studio Repository

Sesame Simulator

The Sesame Simulator, created by Jay Li, is a Rust-based 3D simulation environment for testing Sesame's movements and kinematics in a virtual space. It features: * Physics-based Simulation: Test walking and balance without hardware. * Web-based Interface: Run the simulator directly in your browser. * URDF Integration: Accurate modeling of Sesame's physical properties.

> Go to Sesame Simulator

Sesame ML Tools

ML tools Created by Luke Hollis for Sesame, world model for online planning, policy eval, and RL finetuning. * Robot description: Self-contained printable STLs, CAD-derived MJCF, validated URDF, camera/IMU/foot frames, primitive contact geometry, masses and actuator limits * Tasks: Stand, challenged recovery, commanded locomotion, and goal/obstacle navigation * Policy evaluation: Multi-seed JSON/CSV reports, strict task success, domain parameters, tracking/fall/collision/energy/latency metrics, MP4/GIF and onboard/split views

> Go to Sesame ML Repo

Sesame Companion App

The Sesame Companion App is a Python-based application that enables advanced control and interaction with your robot over your local network. It leverages the new JSON API and network mode features to provide: * Voice Assistant Integration: Control Sesame with voice commands and see real-time emotional expressions. * Remote Control: Command your robot from anywhere on your local network. * Face Control: Change expressions dynamically based on conversation or context. * API Examples: Reference implementation for building your own integrations.

The Companion App works with robots running the latest firmware with network mode enabled.

> Go to Sesame Companion App Repository

Firmware

The ESP32 firmware (sesame-firmware-main.ino) handles the kinematics, face display, and WiFi control interface. * Web UI: Control the robot from your phone via the built-in Access Point. * Custom Faces: Add your own bitmaps (guide in firmware docs).

> Go to Firmware Docs

Contributing

This robot is a platform for building new features, cosmetics, tools, and ideas. Since the current firmware is a basic implementation, pull requests are very welcome for: * Kinematics improvements * New animations * Improved Web UI/UX * Sensor integration (Ultrasonic, Gyro, etc.)

I would also love to see forks of this project with new hardware, software, faces, etc. Be sure to send me a message if you end up building one, and I might feature you on my website or channel!

Created by Dorian Todd. Need help with your Sesame Robot? Send me a message on Discord, my username is "starphee"