AI Kit AI with Hailo 8L for Raspberry Pi 5 - PCIe

AI Kit for Pi 5 based on Hailo 8L module - 13 TOPS power

The Raspberry Pi AI Kit bundles is a Hailo 8L AI accelerator sitted on the Raspberry Pi M.2 HAT+.
Thanks to the Hailo 8L AI, this setup provides an accessible, cost-effective and power-efficient way to integrate high-performance AI with the Raspberry Pi 5.
This kit allows to explore applications like human identification & posture, object identification, process control, smart security, home automation and robotics.

Check out Raspberry Pi's news article for more information

The AI module offers 13 TOPS of processing, 13 tera-operations per second on the neural network inference accelerator. Thanks to the PCIe communication channel, the data can be transfert back and forward at efficient speed.

The AI module can computation results over 13 frames/second with still the Pi 5 Cores freely available for your own application.
A CPU based neuronal network inference will run the 4 Cores of Pi 5 at 100% to compute 2 or 3 frames/second (and no ressources left for your own application).

Many AI application will associate the computation power with a camera (or two cameras for stereoscopic view).
Many AI pre-trained models are availables to take advantage of the neuronal computation power. Please note that AI module is not powerful enough to create and train your own model.

The up-to-date Raspberry Pi OS image for Pi 5 will automatically detects the Hailo module and makes the NPU (Neuronal Processing Unit) available for AI computing tasks. The built-in rpicam-apps camera applications in Raspberry Pi OS natively support the AI module, automatically using the NPU to run compatible post-processing tasks.

The AI kit provides an efficient scheduling of the accelerator hardware allowing to run multiple neural networks on a single camera, OR single/multiple neural networks with two cameras concurrently (source)

Need for using the GPIO with the AI Kit? please check our GPIO Replicator that can be soldered onto the kit stacking header before placing the AI kit.

Note: This product requires a Raspberry-Pi 5 and a camera to work.

What can we do with the kit ?

What can be done with the NPU depends on the available training models.
Fortunately, Hailo does provides lots of pre-trained public models on its hailo_model_zoo repository and Raspberry-Pi's AI Kit documentation will guide you throught the required steps.

The full list for Hailo 8L is available in this sub-page of the repository, some of them are documented in the Hailo Examples for Raspberry-Pi 5.

Here some demonstration video:

• Object Recognition in Traffic (Youtube)
• Object Recognition in Office (Youtube)
• Pose/posture Estimation (Youtube)
• Subject Segmentation (Youtube)

Jeff Geerling also published a very nice Pi AI video about theHailo 8L kit for Raspberry-Pi 5.

Technical details

• For Raspberry Pi 5
• Hailo 8L accelerator with up to 13 TOPS inferencing performance
• HAT+ format
• Pre-assembled on M.2 HAT+ (ready out of the box)
• Thermal pad between module and HAT (spreads heat across components, improving performance)
• Fully integrated into Raspberry Pi’s camera software stack
• Operating temperature: 0°C to 50°C (ambient)
• Can be fitted on Raspberry Pi 5 + Active Cooler

Content

For each order, you will receive the M.2 Hat+ fitted with the AI module pre-assembled. A thermal pad is fitted between the AI module and the HAT.
The bundle is supplied with a 16mm stacking header and spacers/screws. PCIe Ribbon cable included. See the assembling instruction in the AI Kit documentation.

The M.2 HAT+ can be fitted to a Raspberry Pi 5 with the Raspberry Pi Active Cooler in place.

Please note: Raspberry-Pi and Active Cooler are not included.