View Hardware

 

 

Adapting AI pipelines for edge devices is complex. Initially, it involves customizing models for specific hardware, developing AI pipelines, and optimizing performance within resource constraints. Maintaining these pipelines across diverse edge devices also requires ongoing updates and management efforts to ensure compatibility and performance.

 

Lanner partners with Scailable to deliver an end-to-end Edge AI solution development platform, which includes edge AI appliances, model conversion tooling, an AI-pipeline deployment SDK and a cloud based Edge-AI deployment and management platform.

 

Deploy your AI/ML Pipeline in Minutes

Scailable drastically reduces edge-AI pipeline development time from months to hours. With Scailable, data scientists can independently iterate and optimize AI models, while the platform allows field operations teams to deploy new models across a diverse fleet of edge devices seamlessly, without the need for individual device re-engineering.

 

Hardware Independent Deployment

Scailable can deploy the same AI model across different generation of AI hardware platform. For customer that have already deployed the Lanner’s EAI-I130 (Jetson Xavier) and want to add EAI-I131 (Jetson Orin) to the fleet going forward, Scailable automatically adjusts model versions, eliminating the need for embedded engineering. This streamlines time to market, optimizes hardware selection, and simplifies fleet management.

  

Over-the-Air Remote Management

Customers managing a fleet of edge-AI devices in the field will benefit from Scailable for deploying AI models Over-the-Air as soon as models get updated. Scailable eliminates the need for chip-set specific custom development, and facilitates the maintenance across various generations of devices and AI chip-sets that might have been introduced in the fleet over time.
 

 

Scailable on Lanner’s Edge AI Appliances

Scailable simplifies development and shortens the time to market for building Edge AI pipelines for all of the Lanner Rugged Edge AI Appliances powered by NVIDIA, Hailo, and Intel AI architectures. This facilitates migrating AI models from one architecture to another.

 

NVIDIA Edge AI Appliances:

  

Hailo Edge AI Appliances:

  

Non-accelerated Edge AI Appliances:

 

 

With native support for NVIDIA Jetson and GPU architectures, the Scailable platform optimizes and deploys AI models across devices such as the EAI-i130, EAI-i131, LEC-2290E, or any Edge AI server featuring an NVIDIA card.   Scailable supports parallel AI processing across available Hailo AI processors on Lanner’s Falcon H8 card, enabling seamless high-performance AI application development on devices like the LEC-2290H, LEC-7242H, IIoT-I530H, or any Edge AI servers equipped with a Falcon H8 card.   For non-accelerated applications, Scailable supports highly efficient AI inference on native Intel architectures, allowing a diverse range of AI models to run smoothly on devices like the IIoT-i530, IIoT-i531, and any Edge AI server with an Intel CPU running Linux.

 

Ready to Deploy?

 

Request a Demo