Category Archives: Big Data

Recap: Google Environmental Sensor Board and IoT Core (of the Google Cloud Platform)

Google recently released the Coral Environmental Sensor Board ( https://blog.hackster.io/google-releases-coral-environmental-sensor-board-for-the-raspberry-pi-d63245510fbb ), a board with a light sensor, barometric pressure sensor, and humidity/temp sensor, as well as connections for UART, I2C, and PWM. Its further goal is to allow you to securely link your projects and collect, analyze, and process the sensor data using Google’s lineup of various tools. The sensors and connectors on Environmental Sensor Board are interesting but what piqued the interest for the event was Hackster’s article stating:

Google states that the board includes a single key (private, public, and certificate) that will enable communication with the Cloud IoT Core right out of the box.

I was unable to locate (and did not make any significant effort to) Google making that assertion but Google’s Getting Started guide states:

 It includes a secure cryptoprocessor with Google keys to enable connectivity with Google Cloud IoT Core services, allowing you to securely connect

By the time of this posting, I was not able to connect to IoT Core out of the box nor confirm an onboard cryptoprocessor. I was successful using the getting started guide to display sensor values locally on the Environmental Board’s OLED screen. A post on the Coral support repository outlines steps to push data from the Environmental Board to IoT Core. Note that those steps include manually creating a key pair and that the steps were based at least in part with Coral support. Also note that the Environmental Board API doesn’t reference a cryptoprocessor or secure communication. The Coral CloudIot Core API lists a publish_message method which could employ the out of the box secure communications, although the question would remain how to later ingest that published message. No response from direct email to Coral support was received by the time of this post. Unfortunately I agree with the support repository post:

the ‘getting started’ doc for this board just wasn’t deep or broad enough

For the event, we used a slightly modified blend of the IoT Core Quickstart, where attendees pushed data from the Pi to IoT core and viewed the pushed data.

 

 

Sensor Data Analytics Acceleration with Apache Edgent – Notes and Recap

20160909_120015
September Forum

We started the discussion with TexNet,a seismic monitoring program operated by the Bureau of Economic Geology at University of Texas. Additional seismometers are scheduled to be deployed this month, some apparently in Irving and North Dallas. We discussed the system topology.

The event continued with keynote speaker Brandon Swink, an IT Architect working with Apache Edgent, an open source community for accelerating sensor data analytics at the edge.

The presentation (PDF) started with an overview of edge processing and some of the advantages such as reduced communication cost and improved reaction times compared to high latency or severed connections.

Brandon then gave an overview of the current Java development environment incorporating use of existing, publicly available libraries in the ecosystem for faster development of solutions. The demonstration simulated a smart sprinkler system with a Raspberry Pi acting as the edge computer and data simulating soil moisture readings.

Getting Started with Apache Edgent
Apache Quarks on Raspberry Pi with Streaming Analytics
Connect your Raspberry Pi sensors to Watson IoT Platform with Apache Quarks