Real-time data that would be accessible from your geospatial software solution, for traffic, weather or satellite imagery, are one of those allusive promises that you keep hearing about but don't expect to happen any time soon. But at the Esri User Conference in San Diego this week four companies touted their availability from the ArcGIS platform.
DigitalGlobe’s Global Basemap, FirstLook, and Multispectral Premium Services are now integrated with ArcGIS. The FirstLook service according to the company will deliver "near real-time post-event imagery as well as pre-event imagery for comparison. How soon can you get the imagery? We're talking hours but it comes with a premium price as the name implies.
TomTom also announced a new integration service with ArcGIS, providing access to the company's real-time and historical traffic solutions. According to TomTom, "historical data products are based on the largest car centric database of more than six trillion anonymously-collected data points, providing the ability to predict driving behavior across the road network." The obvious applications are for routing as well as predictive analysis for fleet management.
TomTom's competitor, HERE, formerly Nokia Location and Commerce (formerly NAVTEQ), also announced that it was bringing real-time traffic to the Esri platform. It's historical traffic data supplement its real-time data so that as fleets are routed using real-time updates, routing and navigation can be supported by the historical traffic information where real-time data is not readily available. As such, routing algorithms can leverage historical traffic data within urban areas for non-highway or arterial routes.
Finally, AccuWeather is supplying a premium service to ArcGIS users with 200+ metrics of near real time data. These data represent hourly weather data unique to a specific location. AccuWeaather has a content-rich data library going back to 1950 with predicted weather patterns extended forward by two years.
But what are the use cases for real-time data that you would use on a regular basis and that you are willing to pay for? Certainly DigitalGlobe's FirstLook would be essential for disaster response and recovery as well as terrorist attacks. The issue is whether you can get the data soon enough or would UAV-acquired data be a quicker option provided you have clearance to fly.
The use case for real-time traffic is easy because for large fleets, time is money, literally. But the case in industries other than transportation has yet to be made. Historical traffic is sufficient for uses in retail, insurance or urban planning, at least today. Modeling scenarios using historical traffic has great uses but real-time might not have the return on investment yet.
Weather is another story. Paul Raymond from AccuWeather made a compelling argument for Weather-triggered Marketing (WTM). Weather impacts drive-times which might impact daily retail sales. Severe weather impacts certain risk factors associated with driving, flooding, and fire which is of immediate interest to insurance companies to allocate staff and insurance policy rates.
Using historical weather data, Raymond's use cases were some of the most innovative I had heard. He mentioned that one retailer was forecasting SKU demand one year in advance. Others were using weather data for revising store merchandising allocation. Fazolis, a quick service restaurant (QSR), noted that weather impacted labor needs because demand was higher during certain weather patterns. Managers were using weather data so that they could plan staffing more effectively. Another QSR that also sold fuel noticed that there was consistency in selling fuel but inconsistency in selling soft drinks. They used weather data to better understand how it could lead to higher sales.
Raymond also noted that the U.S. saw nearly $20 Billion in weather impacts, 10,000 fatalities and 20,000 injuries due to weather over the last year. "Weather extremes are the new normal," he said.
So, there are certainly times when real-time data is critical. But we may simply be at the early stage of thinking about all the possible uses for which we may need them. It may be a case of "we didn't know we needed it until we had it." It's early days for real-time data and real-time feeds such as those from sensor networks. But in time more use cases will be presented and cases will be made for justifying the additional cost of acquiring real-time data.