Ofer Dekel, principal researcher in the Machine Learning and Optimization group at Microsoft’s research lab, wanted to stop squirrels from eating flower bulbs in his garden and seeds from his bird feeder. So he trained a computer-vision model using Raspberry PI to detect the pesky creatures in his backyard, and trigger his sprinkler system to go on each time they go for the flowers and seeds.
Dekel's backyard project fits with Microsoft's latest paradigm shift into intelligent systems and technology supporting the Bing search engine, but also embedded into the tiniest computing systems.
Microsoft CEO Satya Nadella, during his keynote at the company's Build 2017 conference earlier this year, said by the year 2020, some estimate there will be 25 billion intelligent devices. "We’re moving from what is today’s mobile-first, cloud-first world to a new world that is going to be made up of an intelligent cloud and intelligent edge," he said.
With the ability to collect and analyze data, these devices will be embedded in clothes, scattered around homes and offices, and deployed to perform tasks such as anomaly detection and predictive maintenance everywhere from car engines and elevators to operating rooms and oil rigs, according to Microsoft.
Microsoft doesn't make this point, but in my opinion this data collected from billions of sensors will eventually be used as triggers for digital advertisements, including search.
With a goal to embed artificial intelligence into tiny devices, Dekel's group is working on making these small, simple devices intelligent and secure, even when they're not connected to the Internet of things (IoT).They are taking several approaches using neural networks. You can read more about the project here.