GIS-ALAS Advisory Committee Meeting

Wednesday, May 14, 1997

8:00-10:00am

The purpose of this meeting was to provide updates on the status of the GIS-ALAS project and to address issues related to future tasks.

Update: Software and data distribution options

A meeting was held in April at CTRE to discuss the ways data should be distributed for GIS-ALAS and other GIS-related projects being conducted at CTRE. It was decided that attention should be focused on three GIS vendors: ESRI (ArcView, Arc/Info, Internet Map Server, Map Objects), MapInfo Corporation (MapInfo Professional, MapBasic, ProServer, MapX), and Caliper Corporation (Maptitude, GIS-DK, Interactive Web Mapper). In addition, two distribution options are being considered (CD and the Web). Prototype GIS-ALAS applications will be written on all three platforms. An evaluation matrix (see list of criteria) will be used to help assess the software alternatives and select the best option. Full-scale development will begin immediately after this.

Development of prototypes

A handout was provided (see prototype diagram) which outlines recent activities in generating the accident coverages which are essential in the development of both the prototypes and the full-scale effort. To briefly explain the diagram, on the spatial side, DOT node cartography has been transferred from dgn design files to dxf (a generic graphics format), and then to MapInfo. On the tabular side, the PC-ALAS records (A tables, B tables, and C tables) have been transferred from mpt format to MapInfo tables. The spatial and tabular files are then joined in MapInfo. The final product is a coverage of accident point locations, with tabular data attached and accessible by pointing and clicking. From here, it is a simple matter to convert the spatial and tabular files into Maptitude and ArcView. A coverage of accident locations inor Dallas County, for five years of data, was done in 30 minutes. Doing this for 10 years, for the entire state, should take roughly a week. If the DOT cartography were enhanced, the programs would simply be run again to update the accident locations.

Error checking

The final product of the above process turns out to be a good tool for checking errors. Errors can arise from several sources, such as: missing roads in the DOT cartography (e.g., in recently developed areas), which would make accidents seem to appear on non-existent roadways; and incorrect reporting or coding of "from-nodes" and "to-nodes" in the accident file, which would create new, fictional roadway segments. Accidents can also appear to be mislocated if they occurred on an excessively curved segment of road (the algorithm used to locate accidents uses a straight line distance between the from-node and the to-node). If desired, the cause of the errors can be determined and the cartography and/or accident records can be updated (although this is outside the scope of the current project). The project team is considering possible solutions to the "curved line problem," which may be addressed in Year 2 of the project.

Demonstration of the prototype and error checking

A demonstration was provided, in all three software packages, to show the Advisory Committee the coverages that were created for Dallas County, and to illustrate the data access and error-checking functionality of the current prototype.

Integration with collision diagram software

Duane Smith and Jeff Gerken provided an update on the commercial collision diagram software currently being considered. Two options were found, and one, Intersection Magic, was selected. This can be linked with GIS, but it is not yet integrated). The plan is to design GIS-ALAS and Intersection Magic to work with each other. For instance, an intersection would be chosen in GIS-ALAS, then a button pressed on the toolbar would start up Intersection Magic, which would then provide a collision diagram. (The information would mainly flow from GIS-ALAS to Intersection Magic, and not in the other other direction.) The project team and advisory committee discussed specific issues related to how this would be done, such as what data would need to be transferred from GIS-ALAS to Intersection Magic. In particular, there was a concern to avoid inefficient file structures, duplicate databases, and orphan databases, especially since ultimately GIS-ALAS must also be integrated with MARS, OIM, and other efforts. A demonstration of Intersection Magic was provided. A representative of the vendor will be visiting CTRE to discuss the integration of Intersection Magic with GIS-ALAS. The vendor will also be asked if there is a version of the product that can be run on the Web.

Progress Report

The project team distributed a report on the progress of the project through April (see progress report). The report summarized progress by subtask and provided an assessment of current status ("percent complete"). Some tasks scheduled for completion earlier in the year need to be finished up, and other tasks scheduled for later in the year are being addressed now. Overall, though, the project is generally on track, in terms of both progress and finances.

Future activities

The project team will do benchtesting for the prototype, assessing such things as speed, size of database, and the other criteria in the evaluation matrix (see above). The team has not yet found anything that, from a technical perspective, would preclude the use of any specific package, but further benchtesting is necessary.

At some point, the project team and advisory committee need to assess how the data would be divided for distribution (statewide, RPA, transportation center, highway system, etc.). The final answer will depend both on technical issues and the needs of the end users.

Several aspects of Year 2 activities were discussed, including the role of efforts at the University of Iowa and the University of Northern Iowa. (Steve McGregor of UNI stopped by during lunch to discuss his ideas.) The committee members agreed to meet again soon to discuss the second year of the project. A draft proposal for Year 2 will be prepared for discussion prior to this meeting.