The quarterly Kansas City Arc Info Users Group (KCAIUG) meeting yesterday was hosted by Stantec in Overland Park, KS. It was encouraging and enlightening, and I took quite a few notes. There were 2 presentation parts, which were equally relevant to some clients work I am involved with.
KC Mapping presented first, with discussion towards planning and development best practices. They emphasized the need to follow appropriate planning and development methods to ensure needs are met and addressed without complication. This is more than critical in GIS applications and analysis, where there are so many moving parts and dependencies. In short, a lack of planning or adherence to any sort of development method, is like owning a home without insurance; you may get away with it for a while, but eventually it is going to cost you. They concluded that no one development is better than the other, and that a hybrid Agile/Waterfall approach always seemed to work best, with some situations requiring more emphasis than the other.
The US Army Corps of Engineers (USACE) followed up re-emphasizing the need for appropriate planning, after a long effort of consolidating systems and processes they had in place, which overlapped with those of Federal Emergency Management Agency (FEMA).
Years ago, FEMA recognized the usefulness of USACE data, which they began replicating across numerous databases for their own use, in their own environment. Similar and sometimes exact copies of applications, systems, databases, and processes within the organizations, were leading to massive expenditures on storage and the personnel needed to maintain it all. The lower ranks within FEMA and USACE recognized that they were unnecessarily duplicating efforts within the USACE National Levee Database (NLD) and the FEMA Midterm Levee Inventory (MLI), and quietly initiated a bit of inter-agency collaboration to see how they might relieve some of the burden this short-sighted approach was creating.
The large-scale destruction witnessed from hurricanes in 2007, and the associated awkward responses from ranking officials within both groups, prompted Congress to mandate collaboration between the two organizations. The door was opened, and the stage set for a project that would integrate MLI with NLD. The new integrated backend and dynamic data model eliminated the error-prone and disconnected situation replication and non-cooperation had produced, enabling consolidated applications, systems, and storage, benefiting both organizations on all levels.
Currently, they are working on (and demonstrated) a web application that will be consumed by individuals in the office, in the field, and the general public. It will provide for greater efficiency of systems, higher data accuracy, further reducing workloads. The solution is being produced through a mash-up of open-source solutions, such as MapServer and OpenLayers2, combined with the recognized solutions of Java and Oracle Spatial, providing for a more flexible solution that outperforms a straight-line ESRI solution. This final part is a prevailing theme that I have recognized in most recent approaches to GIS applications; especially, web applications.
In summary… always plan and follow some sort of method, collaborate against the odds, and don’t put all your eggs in one basket.
