ZEDEDA and Edge Impulse Bring ML to the Edge

Edgecompute1

By: Mary Jander


Edge management and orchestration provider ZEDEDA has announced an integration with Edge Impulse, a startup specializing in machine learning (ML) model development for edge applications. The deal could significantly boost the accuracy and timeliness of edge deployments.

Consider, for instance, a smart building app that deploys edge sensors to monitor light fixtures. An application is created to recognize the presence of lights, but it doesn’t differentiate among the kinds of lights—fluorescent or otherwise. Edge Impulse helps a data scientist at the smart building company to create a model that recognizes all kinds of lights to monitor and control their status. ZEDEDA is the vehicle that delivers the model securely as an application to the edge and orchestrates a feedback loop for further data training.

ZEDEDA Responds to Demand

ZEDEDA and Edge Impulse have hooked up in response to market demand. Said Ouissal, CEO and founder of ZEDEDA, described it this way in the press release:

“We’re witnessing firsthand the explosive growth of AI and data-driven initiatives throughout our customer base and have seen the challenges inherent in delivering and iterating these projects within remote and distributed environments. Edge Impulse’s leadership in delivering edge AI models is a great complement to ZEDEDA’s expertise in orchestrating and securing edge applications and devices at scale, and I’m delighted to offer our customers a comprehensive solution to streamline and accelerate their AI transformation efforts.”

ZEDEDA’s presence at the industrial edge is well known. Its software is the basis for edge software from Rockwell Automation, which is also an investor in ZEDEDA. Another investor, Emerson, integrated ZEDEDA’s technology into its DeltaV industrial automation system. While it’s not known whether these vendors will adopt the Edge Impulse solution to work with ZEDEDA, it’s certainly an idea with intriguing potential.

Edge Impulse Hops on ZEDEDA’s Edge

Edge Impulse was founded in 2019 by CEO Zach Shelby (ex-Arm) and CTO Jan Jongboom (ex-Arm and Telenor). It’s garnered $54 million in funding from Coatue, Fika Ventures, Acrew, and Canaan. The company has roughly 70 employees and is based in San Jose, Calif.

Edge Impulse has made a name for itself by offering a kind of studio platform for developers and data scientists to create models that bring machine learning to edge applications. It offers a series of different kinds of model types ready for adaptation to a variety of ML tasks—e.g., detecting and classifying objects or picking out anomalous measurements. Edge Impulse also uses models created on NVIDIA’s Omniverse and with that vendor’s TAO developer framework for creating computer vision models. Via a sizable and impressive ecosystem of partners, Edge Impulse has enabled applications that range from counting pallets in factories to monitoring air quality in print shops.

ZEDEDA brings the infrastructure for delivering these models to edge devices. Edge Impulse models are available in the ZEDEDA Marketplace, from which other applications can be purchased to run on the ZEDEDA platform, which is based on its open-source EVE operating system. ZEDEDA can provide a means of distributing and managing edge apps across thousands of nodes, with zero-trust security.

ZEDEDA's architecture. Source: ZEDEDA

Futuriom Take: ZEDEDA’s integration with Edge Impulse should add momentum to the growing movement to bring AI to edge applications.