Safer Winter Driving in Iowa with FME, ArcGIS Online and Plowcam
Summer is here! And so it seems appropriate to talk about snow. Snow, and winter driving. If you’re from a northerly place like me, one of your early driving lessons may have included your father telling you to stomp hard on the brakes along a stretch of ice-covered country road to “see how slippery it is”, and then hollering “steer INTO the skid!” as you careened sideways towards a large and very stationary pine tree. Which rather counter-intuitively works like a charm. And throughout those long months of winter every year after, apart from that one incident when I found out that not all snow banks have soft fluffy insides, I’ve managed to keep my vehicles more or less pointed in the right direction while cruising down wintery roads. Of course, the massive efforts undertaken by transportation agencies with their fleets of snowplows and equipment had more than a little to do with that too.
Eric Abrams, Geospatial Manager at the Iowa Department of Transportation is intimately familiar with the technology challenges involved in maintaining winter roads. They have more than 900 plows, 9,400 miles of road, typically use 15 million gallons of brine in a season, and in 2013, they put down 120,000 tons of salt on 45 inches of snowfall. Efficiency and information is paramount – a 10% reduction in salt usage can save the department in the area of $1.4 million. Automatic Vehicle Location (AVL) yields a $6.40 ROI for every dollar spent. So as you can imagine, their spatial information systems are big part of the operational picture – and so is FME.
As you can see in this workflow diagram (click on it for a full-size view), Oracle Spatial sits at the center of a flurry of data activity. Each vehicle has been outfitted with a suite of sensors, including dash-mounted iPhones that are set to automatically take pictures whenever the vehicle is moving.. The sensor data is uploaded to Amazon web services, where it is retrieved once per minute to populate the Oracle database. As well, every 15 minutes, FME does a complete data retrieval for aggregation and statistical analysis on materials usage and conditions – for internal usage and monitoring. The photos are sent directly to Oracle using an application developed in-house.
The public-facing portal that the Iowa Department of Transportation has launched is in ArcGIS Online (AGOL). Once a minute, FME retrieves the active plow data from Oracle, and creates a KML file that is uploaded to Windows Azure. AGOL access that KML to build the online representation, refreshed at regular intervals. With the KML files residing in Azure (rather than sending the data directly to AGOL), they are also available for other applications and users to access as needed.
Producing the KML is quite straightforward, as you can see in this workspace. The AVL data is read in, as well as URLs for the pictures and thumbnails. Some date formatting happens, and then the KMLPropertySetter and KMLStyler handle assigning icons and formatting balloons. The KML file is written, and a SystemCaller then uses AZCOPY to move the output data on to Azure where AGOL will retrieve it upon refresh.
Since launching the public site, usage of the data has snowballed. Individuals can visually check conditions before setting out on the road, making for safer driving. Media and weather outlets are taking advantage of this near-real-time situational awareness too, tweeting out plow images to their audiences.
“During the development of Iowa DOT’s AVL system we tried to leverage existing tools and technology at the Department,” says Eric. “FME Server is one of those tools that allowed us to offload processing from our Oracle databases and expedited our public map rollout by avoiding developing code to do the same job. “
And so, between the benefits realized by the DOT in their own internal operations and the potential for safer winter driving by the public, we’d have to say that that’s a whole blizzard of upside. And we’re happy that FME helps make it happen!