ONE Tech Launches First Edge AI that Embeds and Trains AI Models Directly on MCU’s


Reading time ( words)

ONE Tech, a global leader in edge AI technology, has announced the launch of the MicroAI™ Atom Software Development Kit (SDK), which is now publicly available through the ONE Tech developer portal. This SDK download allows users to implement ONE Tech’s machine learning platform—MicroAI™ Atom—directly onto microcontroller unit (MCU)-based hardware.

The MicroAI™ Atom SDK brings intelligence to the endpoint—the source of data—which accelerates machine learning model formation and adoption while lowering overall cost of deploying AI-driven solutions.

The MicroAI™ Atom SDK enables uses cases for assets including:

  • Connected devices (IoT devices, phones, handheld computers)
  • Industrial machinery (robotic arms, welders, material handlers, motors)
  • Utility infrastructure (smart meters, transformers, water treatment facilities)
  • Household appliances (washers, dryers, refrigerators)
  • Automotive (vehicles, heavy equipment)

In a manufacturing use case, for example, there is a critical need to bring intelligence to manufacturing facilities, but existing methods require an overhaul of hardware—an expensive proposition. With the MicroAI™ Atom SDK, these capabilities are available in a low-compute MCU, allowing existing hardware to support machine learning and predictive maintenance to these facilities quickly and cost-effectively.

“This SDK for MicroAI™ Atom is one of a kind, greatly improving time to market with asset performance management and predictive maintenance use cases while lowering the overall cost of deployment via elimination of expensive hardware needed to run and train the model,” said ONE Tech CEO Yasser Khan. “Our team of engineers brought this technology to life and created the framework for developers across the globe to enable intelligence at the asset.”

Share

Print


Suggested Items

Ocean of Things Aims to Expand Maritime Awareness across Open Seas

12/12/2017 | DARPA
The internet of things connects an ever-growing number of smart devices for up-to-the-minute monitoring and tracking of many common events. Head out to most parts of the open ocean, however, and no such capability exists for real-time monitoring of maritime activity.

Today’s MilAero Options: Outsourcing—‘Everybody’s Doing it’ Not so True Today

06/27/2016 | Marc Carter
There was a time, not so many decades ago, when that most commonly-stated mantra (“lower labor costs”) behind offshoring printed circuit fab (and some assembly) operations, still had some case-by-case validity.

How a NASA Team Turned a Smartphone into a Satellite Business

02/19/2016 | NASA
Satellites aren’t small or cheap. The Solar Dynamics Observatory launched by NASA in 2010 weighs about 6,800 pounds and cost $850 million to build and put into orbit. Even the satellites built under NASA’s Discovery Program, aimed at encouraging development of low-cost spacecraft, still have price tags beyond the reach of smaller companies or research organizations.



Copyright © 2021 I-Connect007. All rights reserved.