Part 1: Explanation of the Sample and Template GDB
The Water Utility Templates provide sample datasets and also a template Geodatabase for your use. There are 4 main parts to the Template Geodatabase:
ReferenceData – Landbase data typically acquired from other organizations
WaterDistribution – Water Utility Network Assets
PlanningAndOperations – Long Range Planning and Utility Administrative/Engineering Area Boundaries
Field Operations – Feature classes include redlines/markups and workorders assigned to field crews
We would like to once again thank the Fort Pierce Utilities Authority and St Lucie County, Florida for allowing us to include a sample of their data in the templates. You should be aware that we took a copy of the data at a point in time and built additional content on top of those datasets. As a result, the data in the templates has been significantly altered from the original database and also does not reflect real conditions in Fort Pierce Florida. That will be obvious to many readers of this blog, but we want to clarify that for people who are just getting started.
The Sample.gdb is quite similar to the Template.mdb, but it has more content in the ReferenceData feature dataset. The reason for the difference is that most water utilities do not control the landbase data they use, and we expect that (like Fort Pierce) most of the data loading from partners will essentially be a copy and paste into the target geodatabase. As a result, we only included a few ReferenceData feature classes in the Template.mdb, and depending on your system it is ok to remove those feature classes if you have something different to use.
Below is a detailed explanation of each Feature Dataset in the Sample/Template geodatabase.
This feature dataset contains the water network feature classes. ESRI has built template water network data models for about 10 years and most water utility projects have used those designs as a starting point for implementation. This design should look familiar to most longtime users, but there are some significant design changes here that are described briefly later in this document.
To load data into these feature classes, you should first drop the geometric network. Data loading tools will run faster without the network, and you will also be able to use all of the data loading tools available in ArcGIS. Once you have all of the data loaded you can rebuild the network using the wizard in ArcCatalog.
The loading process is a bit different than the ReferenceData because you have a target database design to load into. Of course it is ok to use a different design or to use the design you already have, but that will mean more work if you want to use the map documents provided with the templates. You might also want to use this latest design if you have existing data because it improves on previous examples.
Generally speaking you will need to do more than copy/paste your water network features into a target geodatabase. You will likely need to calculate new values, combine or split datasets into new feature classes, and you should also have a plan for doing QA/QC on the results of the data loading process. This can be a time consuming exercise that might be more practical to contract out to an organization that specializes in this type of work, and even for simple examples many weeks of work can be involved in getting existing data loaded and validated.
This feature dataset contains administrative and planning content for water utilities. The reason for putting these features in their own feature dataset is that they are edited on a different cycle than the asset data and they may have different permissions/editors.
The simple part of this feature dataset is the feature classes for EngineeringGrid, Map Sheet boundaries, and other administrative areas for water utilities. This part of the database also contains planned replacement mains – wReplacementProject (Capital Improvement Projects) and also proposed mains – wProposedMain (New Development).
This feature dataset also contains content that will be new to most users. What we built in the Fort Pierce database was a set of reporting layers that allow us to organize and view asset inventory, consumption, asset condition, and other data in the context of operational management units like map sheets, districts, and political jurisdictions.
These feature classes were created by spatially joining/intersecting the water network features to sets of polygons like the EngineeringGrid example shown above, and then further calculations for leaks/mile, consumption, and inventory information were performed.
There are 2 main benefits we see from this work:
1. The ability to provide more granular reporting. For example, consumption and number of valves by map sheet or smaller area rather than company-wide reports.
2. Applications such as the Operations Dashboard require that we prepare data so that rather than users running general queries on the data they can just click a few times and get the information they are looking for. As a general principle, we want to “cook” the answers to the common questions to improve performance and reduce load on web servers.
Last but not least the FieldOperations feature dataset contains a number of feature classes that are used in the Mobile and Dashboard Water Utility Templates. These include leaks, redlines, and workorders. While the Mobile Map example only uses the redlines capability in the database, the design includes the ability to manage workorders and inspection information to/from the field as well.
The one thing people notice right away is that workorders and inspections in this data model are Point feature classes. Yes, we know that most workorder/CMMS systems do not store point features because they are tabular systems, but part of simplifying the data/application for field users is finding the geographic location for work. In addition, ArcGIS Mobile works with simple feature classes so this approach works well from an implementation standpoint.
The content for this feature dataset will be more dynamic, and will likely only contain assigned/current work that needs to be pushed to/from the field. For example, today’s work can be loaded into the mobile feature classes and updates can be pushed to the field. Similarly, as work happens in the field, data can be pushed back into the office through a mobile data service.
Ultimately this approach requires some back-end work to get the data from one or more enterprise systems to the field application, and then from the field application back to the enterprise systems. This is not particularly difficult work, but the data loading and management activities will be different than the other feature datasets in the template geodatabase.
Data Model Notes
The data model available with this template has some changes from those data models that reflect the evolution of implementation models for geodatabases. These changes include:
- Uppercase, short field names to ensure that field names are not altered/truncated when moved between different database platforms (and yes shapefiles are still being used for data exchange)
- Longer, more meaningful alias names on fields
- Descriptions on fields and feature classes (FGDC stored as FGDC metadata)
Feature class aliases are plural but table names are singular. When you add a feature class to ArcMap you will get a layer with the plural name (the convention GIS users prefer), but it also works for database people that insist on singular names for tables.
- LifecycleStatus field replaced by ACTIVEFLAG. Previous designs had proposed, active and abandoned features in the same network feature classes. Over time most utilities have separated proposed and abandoned features into different feature classes so they are not accidentally included in network traces or asset inventory reports. Once we added those feature classes to this design, Lifecycle status just became a flag to indicate if the features are active or not. For example, there are temporary/seasonal services that are only used for part of the year, and there are other assets in the ground that are not active because they are still under construction or are temporarily inactive.
OWNEDBY/MANAGEDBY fields indicate the owner of each asset and also the maintenance responsibility. While most domains/list of permitted values in this data model are strings, these domains are integers. The plan here is that “Company owned” assets that should be counted for inventory purposes have a value > 0 and things that should not be counted have a value < 0. In many cases there are multiple companies involved in county/regional scale systems and this strategy will make asset reporting simpler. Of course in those situations people should extend the list/domain of values for their specific situation.
- You may notice that we also dropped AdministrativeArea and OperatingArea fields because these are better managed as part of the reporting layers we built for the PlanningAndOperations data – which is described in the next section of this document.
In each of the Templates, you will see the sample geodatabase and the template geodatabase in the “Maps and GDBs” folder. If you look in the subfolder “Documentation” there are html documents that provide reports for the geodatabase and map documents.
These documents will help you to understand the details of the geodatabases and maps, and also help you to plan out how to load data into your Geodatabase and make changes to map documents to work with your data.
Most projects will start with a source-target matrix spreadsheet that describes the available data and the target datasets for their new Geodatabase. This is a good place to start and it will help you to assess the suitability of the template design as well as the level of effort required to build your Geodatabase.
Again, this can be a large part of your project, so keep in mind that ESRI and Business Partners are here to help you if you need us. Network with your peers to get their recommendations on who to work with, or email us at ArcGISTeamWater@esri.com and we can get you in touch with someone local to help you to get started.