During the Transportation session we looked at several possible options for transportation profiles to SDTS, including -
The near term approach will be development of an ITS profile to SDTS, based on GDF, with
requirements from US GIS-T groups (Dueker/Butler, NCHRP 20-27(3) ), lead by Oak Ridge
(Gordon, Xiong, Franzese) and Viggen (Latham, Siegel), with support from USGS (I hope) and
others. GDF has some linear referencing, but the distances are measured using the stored
coordinate geometry which will not match real-world distances, or official (table) distances,
needed for GIS-T and state DOT users. The latest version of GDF also supports street addresses.
The long term development will be to include transportation requirements in an object-based profile to SDTS (dynamic schema, incremental & transactional database updates, OGIS links), and to expect ISO reconciliation of GDF and DIGEST (used for Canadian transportation) and other standards under an SDTS "umbrella." But, we decided that we couldn't just wait for these developments in two to three years.
The transportation breakout session was not limited to exchange standards, but included talk of the need for standardized content specifications, or data dictionaries, or feature & attribute & relationship catalogs. The need for common semantics was viewed as at least as important as the need for an SDTS profile for the transportation community. There are currently lots of old and new efforts and dictionaries that relate to transportation, e.g., the FGDC's Transportation Spatial Data Dictionary, the GDF 3.0 catalogs for relationships and features and attributes, NATO's Feature and Attribute Coding Catalog (FACC) for DIGEST and VPF, SDTS Part 2, the ITS dictionary, the FGDC work with Tri-Services GIS/CADD Center for a feature registry, and the Oak Ridge collection of definitions from all of the above. Several in the group though that a transportation profile to SDTS should specify a single content specification, but we hope that the profile can work with multiple data dictionaries if necessary. The development of a transportation profile is seen as a separate task from work on a standard transportation content spec. Both activities are important, but should be separate. [I did chastise the group, including myself, for not spending more time reviewing the FGDC GTS transportation spatial data dictionary.]
This breakout session and others sessions included talk of the need for topology in an SDTS
profile. Semantic or logical topology between features (e.g., overpassing and connecting
relationships) was seen as more important to transfer than graph topology among geometric
primitives, which is not implemented, or implemented differently, by different GIS and desktop
mapping products.
We mentioned the need to consider end user needs. Should data in an SDTS transportation profile be applications ready and available for direct exploitation like NIMA's VPF, or should it be optimized for transfer and use ISO 8211 (like GDF & DIGEST)? The complexity of SDTS must be hidden from end users.
Several in the session wanted to establish a newsgroup as way to continue the discussions. Rather than a new newsgroup, we suggested using the existing newsgroup on the BTS site at
http://www.bts.gov/programs/gis/
Would this be appropriate?
Several people suggested having a focused workshop, similar the earlier LRS workshops in Milwaukee and DC, but with the purpose being to continue these discussions and to make progress on common transportation semantics, data models, and exchange standards, and under USGS sponsorship. I did not make any commitments, but USGS would be interested in participating. I expect that we will be working with Oak Ridge and Viggen soon on an ITS profile to SDTS, and this effort may involve a workshop in Knoxville.
Other Transportation Discussions:
The small-group exchanges between David Arctur (with knowledge of SDTS and OGIS capabilities) and Al Butler, and Simon Lewis (with knowledge of transportation requirements) seemed encouraging. David discussed ways to support attribution of parts of objects or partial chains (i.e., dynamic segmentation) via SDTS and object-oriented approaches. David also presented some dynamic schema ideas for SDTS from the DLG-F activity inside USGS, which (DLG-F) was also mentioned in the new Georgia DOT report from BTS at
http://www.bts.gov/gis/state/visit.html
Al Butler provided some information on the latest request for proposals for the National Cooperative Highway Research Program (NCHRP) 20-27(3) project for an improved, comprehensive, multimodal, stable, linear referencing system (LRS) data model. A description of this 18-month project is available through the Transportation Research Board (TRB) research programs page at
The common LRS data model that results from NCHRP 20-27(3) may provide input to the
transportation or object profiles, and to some related NSDI framework activities.
Simon Lewis and Alan Rao talked with several USGS researchers who are very interested in better ways for conflating TIGER with more spatially accurate data from several state DOT's.
|
U.S. Department of the Interior
|| U.S. Geological Survey 1400 Independence Road, Rolla, MO 65401 For general information call: (573)308-3500 URL: http://mcmcweb.er.usgs.gov/sdts/result/breakout/transbrk/transcomments.html Last modified: Thursday, 13-Nov-1997 12:18:26 EST Maintainer: mcmcweb@usgs.gov Privacy Statement || Disclaimers || FOIA || Accessibility |