Introducing SaraKIT: the open source Raspberry Pi CM4 expansion board that enables advanced voice control and precise motor control. This versatile board boasts three sensitive microphones, enabling sound localization for voice recognition up to 5 meters. It also provides two independent BLDC motor controllers, ensuring quiet, rapid, and precise motor control for various applications, including gimbal motors. SaraKIT further supports a CSI interface with two cameras on a flexible cable, integrates two accelerometers, a gyroscope, and a temperature sensor.

SaraKIT is your powerful solution for building modern, efficient voice-controlled products, robots, home automation systems, and interfacing with smart home or office devices. We’ve developed comprehensive demos showcasing how to utilize SaraKIT with popular platforms such as Amazon Alexa, Home Assistant, ChatGPT, Google Home, and many more.

Abundant User-Focused Documentation

Hardware is only as valuable as the accessible software that complements it. At SaraKIT, we prioritize user-friendly design and ease of programming to ensure a seamless experience. We don’t just offer hardware; we provide a comprehensive solution supported by numerous examples and guides. Our well-documented tutorials cover a wide range of topics, from integrating SaraKIT with ChatGPT, Alexa, and Google Home to advanced home automation features like voice control and facial recognition. Whether you’re a novice eager to learn or an experienced developer looking to innovate, there’s something for everyone.

Motor Control Examples with LEGO bricks

Smartphone-controlled RC Car

SaraKIT simplifies the control of quiet and precise BLDC motors. With just a few lines of code and LEGO bricks, you can swiftly construct a sleek, fast, and silent LEGO car. This car features highly precise, independent 4-wheel drive with two differentials. You can control it via your smartphone. The possibilities extend further by adding a camera for monitoring, implementing line tracking, object recognition, or even voice control. With Raspberry Pi’s capabilities, your imagination sets the limits.

Simplicity, intuitive handling, and ease of programming are important to us. We’ve made programming the motors exceptionally straightforward. For example, to make the wheel travel precisely 25 cm, you only need two lines of code:

#BLDCMotor_DriveMeters(uchar motorId, float centimeters, float speed, uchar torque, float WhellDiameter);
BLDCMotor_DriveMeters(0, 25, 3, 10, 7);  

If you’re keen to embark on your own journey of building your own Pi-controlled car, we invite you to visit our GitHub repository. Delve into our rich collection of source code available in C++, Python, and Delphi. Experiment, customize, and enhance your project based on your own vision.

Self-Balancing LEGO Robot

SaraKIT unlocks new horizons for DIY enthusiasts. Its precise motor control combined with accurate wheel position reading allows you to create a self-balancing robot that’s genuinely hard to topple. This level of control is unattainable with noisy, imprecise servos.

This example demonstrates of the speed and power of BLDC motors:

Precise BLDC Motor Control

Controlling gimbal motors becomes effortless with SaraKIT. You can adjust speed, torque, and angle with ease. Why settle for less when you can harness seamless integrations?

We’ve simplified the program code so that you can achieve what you see in the video above with just a few lines of code:

Pan-tilt Cameras

Control a quiet, precise pan-tilt camera (or even two) using BLDC gimbal motors. Say goodbye to noisy, imprecise, and unreliable servos. Utilize powerful, fast, and nearly silent BLDC motors for precise angle adjustments, capable of control down to a fraction of a degree.

Face Detection, Object Detection, and Tracking

SaraKIT is an easy-to-use face analysis solution for Raspberry Pi 4 CM4, powered by state-of-the-art algorithms based on Google’s MediaPipe. It excels in face detection, face landmark detection, and face mesh processing, optimized for the Raspberry Pi 64-bit platform.

Our pan-tilt camera system, based on SaraKIT, delivers excellent face detection and tracking capabilities, thanks to the quiet, precise BLDC gimbal motors. This innovative solution represents the many ways to leverage SaraKIT in projects related to artificial intelligence and home automation, opening up new possibilities for monitoring, facial recognition, and real-time interactions.

Sound Locator

With three microphones and ZL38063, SaraKIT can analyze sound direction location. The ZL38063 beamformer reports the angle of the most significant sound source for linear/triangular microphone configurations. This video demonstrates SaraKIT spatially tracking a sound source, produced by a phone playing an audible tone.

Stereo Vision (Depth)

The Raspberry Pi CM4 version supports two cameras, a feature lacking in the standard Raspberry Pi. SaraKIT takes full advantage of this capability. You can easily connect two cameras for depth perception mode. In our GitHub repositories, we will provide sample code and ready-to-print camera holder files for two cameras in STL or Blender formats.

Support for ChatGPT, Alexa and Google Home

With three sensitive microphones, SaraKIT can recognize speech from up to 5 meters away. This makes it easy to integrate the device with ChatGPT, Amazon Alexa, Google Home or Home Assistant, and other voice control systems.

An example of this integration can be seen in the demonstration video:

SaraKIT eliminates the need for dedicated devices for ChatGPT, allowing you to create your own voice assistant. What’s more, the version with a pan-tilt camera, equipped with face detection and tracking capabilities, enables device activation with just a glance, eliminating the need for wake words like "Alexa" or "OK Google."

SaraKIT with ChatGPT

With SaraKIT, you can build a ChatGPT-supported device housed in a 3D-printed casing with user face tracking and recognition capabilities.

Say goodbye to constant wake word utterances like "Alexa" or "OK Google". Just look at the camera before issuing a command, and your enhanced assistant will recognize your intent and know you’re addressing it, making interactions more natural and intuitive, like human conversation.

Furthermore, with the facial recognition feature, the device knows who it’s talking to. By observing its surroundings, it not only hears but also sees and asks questions based on what it perceives around you. With SaraKIT, you can create a ChatGPT Assistant that goes beyond just voice-control, and elevate the capabilities of your assistant to a higher level. Check out the video below, showcasing the potential of SaraKIT combined with ChatGPT. In this video, we used a professional casing for SaraKIT, which is entirely optional. Our pan-tilt enclosure is perfectly suitable for this application.

Natural Conversations with ChatGPT

Home Automation - Voice Control

Elevate your home experience with SaraKIT’s voice control capabilities. Seamlessly integrate SaraKIT with leading voice assistants and command your home devices with just your voice. No more searching for switches or manually adjusting settings. Whether you’re setting the mood with ambient lighting, adjusting room temperatures, or playing your favorite tunes, SaraKIT streamlines the process.

We’re actively developing examples that illustrate the synergy between SaraKIT and the Home Assistant system. These guides aim to provide you with hands-on guidance to harnessing its full potential.

Open Source Flexibility

SaraKIT is not just a product; it’s a community-driven initiative. Open-source at its core, it invites enthusiasts and experts to contribute, modify, and innovate. You can learn more by exploring our open-source files in our Github repository.

Unboxing Demo

We’ve taken care to ensure that SaraKIT it is well-protected during shipment. SaraKIT consists of three elements: the Raspberry Pi Compute Module 4 Carrier board, a dual CSI camera flex cable (17 cm), and a Raspberry Pi camera standard CSI connection adapter. The J2 and J3 connectors are standard CSI connectors, allowing compatibility with Raspberry PI-compatible cameras. Get a sneak peak with our unboxing video.

Features & Specifications

• Carrier board for Raspberry Pi Compute Module 4 (CM4) with future CM5 compatibility.
• Compact design, slightly larger than a credit card (⌀ 9 cm).
• Advanced voice control with audio based on the ZL38063.
• Three sensitive microphones via SPH0655.
• Sound localization for voice recognition up to 5 m away.
• High sensitivity with an SNR of -37 dB ±1 dB @ 94 dB SPL.
• Amplified stereo output: 2 x 6 W, 4Ω.
• Two three-phase BLDC motor drivers DRV8313, supporting a maximum of 65 V and 3 A peak.
• Two Encoder inputs (reprogrammable as GPIO).
• 11 GPIO pins for UART, I2C, PWM, and more.
• Dual camera interfaces (CSI).
• Digital Accelerometer LIS3DH (3-Axis, 2g / 4g / 8g / 16g, 16-bit) with I2C/SPI interfaces.
• LSM6DS3TR, providing an always-on 3D accelerometer, 3D gyroscope, and temperature sensor.
• dsPIC33, an embedded, programmable, 16-bit microcontroller with 32 KB of memory.
• Host USB support.

Hardware Overview

Hardware Logic

Power supply

SaraKIT requires a power supply of 12-24 VDC with a minimum of 24 W when using motors in your project. We recommend using a correspondingly larger power supply. For mobile devices, connect to a USB-C PowerBank (PD2.0, PD3.0) with a PD 12 V adapter cable, or via a USB-C cable with a USB-C PD 12 V Trigger Module.

Comparisons

Support & Documentation

• Open Hardware Files
• Software Files
• Getting Started Guide
• Support Forums (coming soon)
• Other Documentation