curb-data-specification
curb-data-specification copied to clipboard
openmobilityfoundation
Data Structure challenge
Our data structure for our meters is inconsistent with CDS standards and that is one of the challenges converting our current data to CDS. For example, our IPS meters data structure is structured in that order: Zone/Area/Sub-area. All of which are inconsistent with CDS
@alkhoury-elias It sounds there's both a semantics issue and a data structure issue here. I think it's worth separating the two so we can determine what, if any changes are needed in the spec.
For the semantics issue, it appears what's referred to as a "Zone" in your IPS data is called an "Area" in CDS and vice-versa. This doesn't seem like it would warrant a change because it's a straight forward translation issue.
For the structural part, CDS defines a 3 level hierarchy: Area/Zone/Space
Whereas your IPS uses a 4 level hierarchy: Zone/Area/Sub-area/Space
It's also important to note that these CDS objects (Area/Zone/Space) must conform to a number of criteria described here and it's unlikely that existing, non-CDS data is going to conform without some changes to your data.
Are there specific use cases that require a 4 level hierarchy or does it make more sense to restructure your IPS data?
Does anyone else have examples where a 4 level hierarchy would be necessary?
I will add we have worked with IPS meters before and we have often stored these 'extra' hierarchies in areas. For example, in Arlington County they grouped sections of paid areas by 'payment zones' (areas where prices would be adjusted in unison), paymentID areas (areas that use the same ID for payment), and block faces (a group of curb zones along a given block). We stored all of these in Curb Areas, and used the 'name' attribute to clarify what category of Curb Area they fell into. This worked nicely to keep the County's grouping standards while abiding by CDS.
At Populus we treat the organization of meters for city purposes separately from CDS. We get the location of a meter from CDS or some other method and then our platform can connect it with any number of other geographic areas for purposes of analysis.
It seems that CDS can cover the IPS data structure hierarchy, and that the semantics of the names is not an issue. So can we close this issue now @alkhoury-elias or do you think there is more discussion needed?
I wanted @rneubauer to weigh on it.
Elias Khoury Division Manager – Parking & Downtown Operations City of San José - Department of Transportation
I'll take a look at it, but I do believe it is covered in the current deployment we have. The only thing to note is that we had to adjust the IPS data to get it to fit. It wouldn't fit in with their standard Schema.
Rick Neubauer Founder and CEO w: www.umojo.com
Closing for now but we can reopen later if needed, or create a new issue.