Safe Software continues to innovate far beyond the extract, transform and load of different spatial data formats. The company takes its role of being an aid to workflows seriously, and has recently added functionality to their FME Server that provides a conduit between sensors and systems. I spoke to Dale Lutz, co-founder & VP Development, on Friday and was fascinated about where these advances take us.
FME Server allows for the scheduling and automation of tasks, such as taking data transformations and pushing them to a web service. Previously, these triggers were activated through a form from a URL, but with the 2012 release, the transformations can be triggered by a stimulus event, such as a URL post from a sensor. This functionality directly enables the Internet of Things, by allowing sensor communications to become actions.
A simplified example of this is to couple data posts from an Energy Monitoring System in your home to be routed to an FME Server where the readings can be automatically aggregated. The sensor readings could also automate an action based upon a given threshold that triggers other events. As an increasing number of sensors are enabled to send messages via e-mail or SMS, the ability of the FME Server to parse the text of an e-mail to read the machine details and apply the readings to existing business rules really opens up all-new capabilities.
Safe has been testing this capability of automating and reacting to near real-time data feeds with a few applications for the iPhone, where all sensors of an iPhone give information regularly, including reporting where you’re going. FME Server can compare this incoming information on the fly, and helps to glue data pieces and feeds together.
In addition to this new capability with sensors, there are also some breakthroughs in the processing of Lidar data, including the ability to handle a billion lidar points and load them into Oracle. As the capture technology continues to push more and more points, particularly with mobile lidar, the ability to handle billions of points, to stitch the points together, and to split based on color and intensity becomes compelling.
Lutz shared a mobile lidar image where a user had used intensity and color splitting to parse out white lines and yellow lines in a roadway. This functionality points to a whole new automated feature extraction that might even be tuned to turn the points into polygons. The latest toolset also allows users to automatically add color to lidar point clouds by associating the points with an orthophoto.
I’m enthusiastic about this company’s continued capability in 3D and real-time data processing, along with all their other tools to improve workflows and aid data analysis. The vision for a more real-time and reactive spatial system is being realized.