Sometimes it’s not enough to be vigilant. Sometimes you need tech on your side. Enter the lowest-cost, most discrete personal monitoring reference designs from AI trailblazer Ambient Scientific!
In another on-booth partner interview from ipXchange at Embedded World 2024, Eamon catches up with GP Singh for an in-person examination of Ambient Scientific’s revolutionary AI technology.
For those hearing about Ambient Scientific for the first time, check out ipXchange’s previous interviews by following the links below, but the pitch is simple: If Artificial Intelligence (AI) can be used to greatly improve our lives, it needs to be affordable in order to be deployed anywhere. And in order to be deployed anywhere, it needs to be able to run at the end point – i.e. without having to connect to the cloud – with extremely low power consumption if you intend it to run on batteries or harvested ambient energy.
And this is what GP brought to the show: several reference designs that illustrate the compact, low-power nature of Ambient Scientific’s GPX-5 and GPX-10 AI processors for always-on AI processing and inferencing at 80 µW or less.
These designs include lockets for personal health and safety monitoring – such as if the user has an asthma attack or is assaulted – as well as weapon-detecting cameras for intelligent security in schools or public buildings.
Best of all, Ambient Scientific has built these chipsets at a price point that enables such products for tens of dollars, making it a no-brainer to adopt these devices should you decide to build a product yourself. This is especially true if a large number of end nodes are required for an installation.
Follow the link to the board page below to learn more about the technology behind this innovation, and if you’re looking to implement AI vision, sound classification, speech recognition, or sensor fusion in long-lasting end products, apply to evaluate this technology today.
Keep designing!
Enjoyed this discussion? Check out our other chats with GP Singh regarding these game-changing AI chipsets:
Want ‘Michelin-Star’ AI? Master the basic ingredients first! Analogue matrix computing at 60 µW