The Internet of Things promises to enable a lot of, well, things. From smart homes to smart cities, and everything in between, tech enthusiasts are well acquainted with the potential of a world populated by connected devices.
Silk Labs, a startup headed by former Mozilla CTO Andreas Gal, believes the key to unlocking that potential lies in more intuitive experiences. Earlier this year, Silk Labs launched a Kickstarter campaign for its Sense camera. Capable of face and voice recognition, gestures, and deep learning, Sense is an example of how next-generation of IoT devices are shifting away from the traditional ‘smartphone-as-a-hub’ model.
The IoT equation
If you boil down IoT to its components, you’d find it’s equal parts software, cloud services, and hardware. And while Silk Labs has cloud computing and software covered, the hardware for Sense is none other than the Qualcomm Snapdragon 650 IP camera reference design and development platform.
Built around the Snapdragon 650 processor, this platform was designed for the next generation of smart cameras. With features such as analytics, advanced imaging, and 64-bit processing, the Snapdragon 650 IP enables Sense to learn from its environment. The camera is engineered to recognize friends, family (and even pets) not only by their faces, but also by their voices. When the camera detects an unknown guest, its app asks owners to identify that guest. Integrating your devices with Bluetooth also means it can tell who’s at home by communicating with other smart devices.
Facial recognition in security cameras isn’t exactly new. The difference is that with Snapdragon 650 IP’s advanced analytics capabilities, Sense can intelligently discern how to use the information it collects. Instead of alerting owners of every visitor, Sense sends notifications to users or video to the cloud only for guests it does not recognize, odd sounds, or unusual motions.
Thinking outside the cloud
The term “smart” devices can be deceiving—the fact that a device can access the Internet doesn’t mean it’s harnessing that connectivity intelligently.
Powered by a 64-bit hexa-core CPU, Sense runs applications locally. Unlike other connected devices that operate completely on the cloud, local computing enables greater privacy through end-to-end encryption. Additionally, only users can decrypt and view video streams, while keys are shared securely based on proximity data via Bluetooth.
That being said, the Snapdragon 650 IP provides comprehensive connectivity through the X8 LTE modem and 802.11a/b/h/n/ac Wi-Fi, as well as other options such as GPS, Ethernet, and Bluetooth.
The Sense camera is expected to be available later this year—but in the meantime, anyone eager to build the next generation of IoT devices has plenty of resources to get started.
- Related content: Guest blog: Silk Labs build their product using development boards and resources from Qualcomm Developer Network
