By Michael Manoochehri, Developer Programs Engineer, Google Cloud PlatformCross-posted with the Google Cloud Platform Blog
After last year's Google I/O conference, the Google Cloud Platform Developer Relations team started to think about how attendees experienced the event. We wanted to help attendees gain more insight about the conference space and the environment itself. Which developer Sandboxes were the busiest? Which were the loudest locations, and which were the best places to take a quick nap? We think about data problems all the time, and this looked like an interesting big data challenge that we could try to solve. So this year, we decided to try to answer our questions with a project that's a bit different, kind of futuristic, and maybe a little crazy.
Since we love open source hardware hacking as much as we love to share open source code, we decided to team up with the O'Reilly Data Sensing Lab to deploy hundreds of Arduino-based environmental sensors at Google I/O 2013. Using software built with the Google Cloud Platform, we'll be collecting and visualizing ambient data about the conference, such as temperature, humidity, air quality, in real time! Altogether, the sensors network will provide over 4,000 continuous data streams over a ZigBee mesh network managed by Device Cloud by Etherios.
In addition, our motes will be able to detect fluctuations in noise level, and some will be attached to footstep counters, to understand collective movement around the conference floor. Of course, since a key goal of Google I/O is to promote innovation in the open, the project's Cloud Platform code, the Arduino hardware designs, and even the data collected, will be open source and available online after the conference.
Google Cloud Platform, which provides the software backend for this project, has a variety of features for building applications that collect and process data from a large number of client devices - without having to spend time managing hardware or infrastructure. Google App Engine Datastore, along with Cloud Endpoints, provides a scalable front end API for collecting data from devices. Google Compute Engine is used to process and analyse data with software tools you may already be familiar with, such as R and Hadoop. Google BigQuery provides fast aggregate analysis of terabyte datasets. Finally, App Engine's web application framework is able to surface interactive visualizations to users.
Networked sensor technology is in the early stages of revolutionizing business logistics, city planning, and consumer products. We are looking forward to sharing the Data Sensing Lab with Google I/O attendees, because we want to show how using open hardware together with the Google Cloud Platform can make this technology accessible to anyone.
With the help of the Google Maps DevRel team, we'll be displaying visualizations of interesting trends on several screens around the conference. Members of the Data Sensing Lab will be on hand in the Google I/O Cloud Sandbox to show off prototypes and talk to attendees about open hardware development. Lead software developer Amy Unruh and Kim Cameron from the Cloud Platform Developer Relations team will talk about how we built the software involved in this project in a talk called "Behind the Data Sensing Lab". In case you aren't able to attend Google I/O 2013, this session will be available online after the conference. Learn more about the Google Cloud Platform on our site, and to dive in to building applications, check out our developer documentation.
Michael Manoochehri is a Developer Programs Engineer supporting the Google Cloud Platform. He is passionate about making cloud computing and data analysis universally accessible and useful.
Posted by Scott Knaster, Editor
