The global availability of mobile technology means that everyone is connected on the go. For businesses to truly penetrate the consumer market in the age of on-demand products and services, they must find a way to make use of mobile data. Every data point has a place – this is where geospatial data analytics comes into play. The ability to analyze geospatial data and build applications that utilize location will separate market domineers from names left behind. If you know when and how connected consumers interact in different places, can harness data from sensors and IoT, and machine-to-machine communication, you can deliver the most efficient, personalized experience.
MemSQL Meetup with Mapbox: Visualize Your World with Geospatial Data
Our next meetup spotlights geospatial data analytics. Join us in SoMa, San Francisco to learn about innovative tools for mastering geospatial analytics and building geo-enabled applications.
Featuring Matt Irwin of Mapbox
Matt will showcase data visualization with the Mapbox platform. Neil Dahlke of MemSQL will follow with a live demonstration of PowerStream and Supercar, two real-time applications built on top of Mapbox.
MemSQL comes equipped with geospatial capabilities. Let’s explore a few examples.
An IoT showcase application, PowerSteam models predictive analytics for global wind farm health. Individual wind turbines in these farms have a number of sensors attached to them, assessing temperature and vibrations. In order to harness this data, we constructed a real-time data pipeline, beginning with Apache Kafka, a distributed message queue, at the front. Sensor data is then streamed into MemSQL via Streamliner, an integrated Apache Spark solution. Add in a Spark MLlib predictive model to determine whether a wind turbine is likely to fail, and serve the data up through MemSQL to a front-end dashboard. In this case, the dashboard is provided by Tableau, and the scrolling map showcasing the predicted health of wind turbines is provided by Mapbox.
Watch MemSQL CTO, Nikita Shamgunov, build PowerStream live below:
Supercar is a geospatial application that analyzes New York taxi rides in real time. The dataset contains location information from actual rides that took place throughout Manhattan. With Supercar, you can run easy SQL queries in MemSQL to analyze the average length of ride, or average cost of ride, for example.
The data science team at Pinterest uses Apache Kafka, Spark Streaming, and MemSQL to ingest tens of thousands of events per second and aggregate that event data, which in this case is pins and repins. Kafka serves as the message queue, Spark provides the transformation tier, and the operational database offers persistence and a serving layer for an application that allows for quick analysis of real-time trending topics by geography.