- With Azure Percept, Microsoft aims to help enterprises deploy and use Azure AI services on devices located at the network edge to support new use cases.
- Despite a swath of capabilities to support use-case development, Microsoft still needs to demonstrate demand for the sort of use cases Azure Percept is designed to enable.
One of the new solutions announced by Microsoft at its recent annual developer conference, Ignite, was Azure Percept, a hardware- and software-based platform to help enterprises deploy and use Azure AI services on devices located at the network edge. Azure Percept includes a development kit, Azure Percept DK, which allows users to develop AI proofs of concept for use cases leveraging vision and/or sound. These could include applications that detect production anomalies in manufacturing scenarios or those that apply shelf or customer analytics to retail environments. The kit features hardware-enabled AI modules for running models at the edge. However, they can also be connected to the Azure cloud, supporting enterprises’ ability to take advantage of hybrid cloud scenarios.
The development kit can be integrated with a camera called Azure Percept Vision or with Azure Percept Audio, a device containing a preconfigured audio processor and a four-microphone linear array, enabling users to apply voice commanding, keyword spotting, and far field speech to local listening devices using Azure Cognitive Services. Azure Percept devices can automatically connect to Azure IoT Hub to enable secure communications between IoT devices and the Azure cloud.
In addition, Azure Percept Studio provides users with a single portal for creating edge AI models and solutions. Users can access a library of pre-built Azure AI models, or they can choose to build custom no-code models of their own. This will benefit customers who do not have a lot of coding experience, guiding them through the process of developing, training, and deploying AI proof-of-concept ideas.
At Ignite, Microsoft also revealed that it is working with third-party silicon and equipment manufacturers to build an ecosystem of intelligent edge devices that will be certified to run on the new Azure Percept platform. The Azure Precept development kit uses the industry-standard 80/20 T-slot framing architecture; this should help facilitate the development of devices that are compatible with the platform. All Azure Percept DK devices are designed with additional built-in security on every device. This will help to protect privacy-sensitive sensors such as cameras, microphones, and inference data, as well as ensuring device authentication and authorization for Azure Percept Studio services.
Microsoft already offers several solutions that enable enterprises to leverage data processing and app development capabilities closer to places where data is generated and collected and to where low-latency applications are consumed. The new Azure Percept offering adds new AI capabilities to the mix and expands the range of potential use cases for Microsoft’s edge technologies. Going forward, the challenge for Microsoft will be to demonstrate, or stimulate, end-user demand for the sort of use cases Azure Percept is designed to support.