On-device artificial intelligence for embedded vision solution – empowered by Hailo AI

In an industrial landscape where logistics and outdoor applications are becoming increasingly autonomous, a robust edge-based software solution is essential. How can one develop an autonomous perception sensor with cutting edge on-device AI capabilities? SICK showcases this by integrating Hailo’s AI processor into its Visionary-B Two vision device, further enhancing embedded software intelligence in the automation industry.

SICK demonstrates how to develop a sensor with AI capabilities directly on the device by integrating the Hailo AI processor into the Visionary-B Two 3D camera.

Centralized vs. embedded approach

As industrial operations move towards full autonomy, the demand for computing power, particularly for AI-based software applications, surges. One common solution is to offload computing tasks to external industrial PCs (iPCs) equipped with GPUs, so called a centralized approach. However, what if AI applications need to run directly on the device? This presents a significant challenge, as increased computing power generates more heat, limiting the device’s operating temperature range—a critical factor in industrial settings. Additionally, the total cost of ownership for on-device solutions must be considered.

Computational tasks are offloaded to external, GPU-equipped industrial PCs (iPCs).

More and more data need to be computed with minimum latency to achieve the next level of automation

Expertise from a high-performance AI leader

This is where Hailo’s unique expertise comes into play. Hailo, a leading manufacturer of high-performance AI processors, designs solutions to run advanced machine learning applications at the edge. This collaboration boosts SICK’s on-device AI capabilities. Hailo’s low-power AI computing fits seamlessly into the on-device architecture of SICK’s sensor autonomous perception portfolio by first showcasing it with vision-based products.

The energy-efficient AI computing from Hailo integrates seamlessly into the on-device architecture of SICK’s Vision portfolio.

Visionary-B Two with on-device programming capabilities

On-device programming with 3D vision camera and AI Accelerator

The 3D camera Visionary-B Two with Hailo-8 AI Accelerator enables native programming and the use of Hailo software tools.

In October 2024, the first product of this collaboration hits the market: a 3D snapshot device featuring the Hailo-8 AI accelerator. The aim of this variant is to allow users to natively program on the device while leveraging the Hailo AI Accelerator and its software tools. Key application areas include logistics and outdoor automation. Software partners of SICK and Hailo are already reaping the benefits of this technology in various pilot projects, developing AI-based applications. Both SICK and HAILO are bringing in innovative software partners from their respective business areas.

The new solution is being showcased live at VISION 2024, the world’s leading trade fair for machine vision. You are welcome to see this innovation live at HAILO booth at Hall 8, 8B25.

Machine vision

Rugged and smart – 3D snapshot solution for harsh outdoor conditions

Visionary-B Two

On-device artificial intelligence for embedded vision solution – empowered by Hailo AI

Centralized vs. embedded approach

Expertise from a high-performance AI leader

On-device programming with 3D vision camera and AI Accelerator

Read more

Do you have any questions?