Introduction to Google’s New Line of Coral Products

One of its most recent and cutting-edge creations is the Coral line of products. This new line of developer and engineer-focused tech is built to make prototyping and exploring revolutionary concepts like machine and deep learning, visual AI, and IoT easier, faster, and more accessible. Read on to learn about these products and how they can make your advanced computing products come to life.

Coral Development Board: An Adaptable Dev Board

The Coral Development Board is a single-board computer (SBC) that can be used for a variety of applications, from edge computing in industrial IoT solutions to machine-learning and artificial-intelligence builds. This versatile and flexible platform features a removable computer module, which makes it ideally suited for both prototyping as well as end-application use.

With the proliferation of products centered on robotics, facial recognition, intelligent control, and other advanced computing systems, the Coral Development Board offers major computing power to keep up with these demanding tasks. It supports TensorFlow Lite, a deep-learning framework used by some of the biggest names in tech, including Google, Nest, Shazam, and more. It also supports Python and C++ and operates using Mendel Linux, which is a derivative of Debian. With GPIO connectivity, it is easy to connect the Coral Development Board to external systems and sensors.

Coral USB Accelerator: A Google AI Chip TPU

The Coral USB Accelerator is an on-board Edge TPU co-processor that packs a ton of power in a tiny package. This incredible product is capable of performing 4 trillion operations per second (TOPS) while using just 0.5 W for each TOPS. With this combination of computing power and efficiency, the Coral USB Accelerator is extremely well-suited for edge applications in which power may be scarce but you can’t compromise on performance.

The Coral USB Accelerator is designed to work with a Debian-based Linux system via the included USB 3.0 Type-C cable. Perhaps the best thing about the Coral USB Accelerator is that it gives users the ability to add high-end machine-learning capabilities without a high-end device like a mainstream PC. In fact, the USB Accelerator can add intelligent system capabilities to the world’s most popular SBC—the Raspberry Pi.

With the Coral USB Accelerator, there is no need to build models from the ground up. Rather, existing TensorFlow Lite models can be compiled to run on the Edge TPU. Finally, the Coral USB Accelerator supports AutoML Vision Edge, making it easy to build and deploy custom image classification models to your device.

Coral 5-MP Camera Module: A Google Machine Learning Camera

The Coral 5-MP Camera Module is custom-built to interface seamlessly with the Coral Development Board. This 5-megapixel camera module is well-suited for facial-recognition, emotion-detection, and object-detection capabilities. The Coral Camera would also be at home in industrial IoT applications such as part and product detection. Other applications could be face-tracking capabilities in a single-person video-shooting platform or a security system that records only when unrecognized people are detected.

On the hardware side, the Coral Camera Module has an omnivision sensor, auto-focus, and auto-exposure control. Coming in at just 25 × 25 mm, the small size of this camera module makes it an excellent choice for embedded applications in which size/weight is a consideration. The Coral Camera Module easily connects to the Coral Development Board via the board’s built-in CSI camera connector, giving you truly “plug-and-play” imaging capabilities. Lastly, the Coral Camera Module is designed with scalability in mind and is built to scale as your manufacturing requirements change.

Coral Environmental Board: Sensors with Cloud Capabilities

One of the biggest impacts of the internet of things has been through its ability to measure not just the data from a device but the environment around the device itself. The Coral Environmental Board is built to add some of the most popular environmental sensors to the Coral Development Board with ease. It includes temperature, light, and humidity sensors, along with four additional Grove connectors to allow for quick expansion. The inclusion of a secure element allows it to work with Google’s Cloud IoT Core. The Coral Environmental Sensor Board is compatible with both the Coral Development Board and Raspberry Pi.

With the included sensors, the Coral Environmental Board offers plug-and-play capability as a remote weather station. It can also use the readings from its accurate slate of sensors to record environmental readings for IoT applications. By combining the Environmental Board with the machine-learning capabilities of the Coral Development Board and the USB Accelerator, a user could design a platform to control machinery more efficiently by learning from the environment.

AIY Vision Kit: Build Image Recognition

If you are looking for an all-in-one solution to start using motion detection, object recognition/avoidance, or image recognition, the AIY Vision Kit is exactly what you need. The AIY Vision Kit is built around the Raspberry Pi Zero—the small but mighty SBC with a small footprint but an incredible amount of computing power. With a built-in TPU, image, face, and object recognition are all within each, and the host of inputs and outputs (GPIO, PWM, and ADC) means that the possibilities are endless.

The AIY Vision Kit utilizes an expansive Python library, which means that programming many tasks is easy and fast, making this kit a great choice for prototyping. From emotion detection and face tracking to lightweight robotic systems with visual machine learning, the range of projects that this powerful kit works with is endless.

Find more information on the new line of Google Coral products, including specifications and accessories, click here. Browse other Google AIY Accessories at Arrow.com.