RE: 2D Building footprint not accurate

[DelageBaza, Christina (DOA)]

Hi Caterina,

I also noticed many of these inconsistencies when the Data Management team was reviewing and providing feedback.  I’ve included Chuck & Aimee as I was not involved in this contract, but they may have more insight and feedback on this deliverable.  Maybe there can be additional post processing through machine learning/AI to remove these features?    

 

Christina Delage Baza, GISP
Data Analyst I
RIDOA - Division of Statewide Planning
235 Promenade Street

2nd Floor, Suite 230

Providence, RI 02908

 

401-222-6481

https://planning.ri.gov/planning-areas/data-center

 

 

 

From: Massidda, Caterina (EMA) <Caterina...@ema.ri.gov>
Sent: Wednesday, November 5, 2025 12:51 PM
To: RIGIS <rigis...@uri.edu>; DelageBaza, Christina (DOA) <christina....@doa.ri.gov>; HoxsieQuinn, Garth (DOA) <Garth.Ho...@doa.ri.gov>
Cc: Washburn, John (EMA) <John.W...@ema.ri.gov>; Guthlein, Thomas (EMA) <Thomas....@ema.ri.gov>
Subject: 2D Building footprint not accurate

 

Hi Christina and RIGIS,

I downloaded the newly published Building layer on RIGIS and, unfortunately, noticed that several of the errors and feedback previously reported were not addressed.

I remember submitting a few hundred comments, along with other members of the Data Management team who participated in the review, which aren’t reflected in the final product. Here are just a few examples of the issues still present:

1. Solar panel farms are still being recognized as buildings.

2. Hedges are identified as buildings.

3. Boats and containers are also being classified as buildings.

 

 

Due to these inaccuracies, RIEMA is unable to effectively utilize this layer. We kindly request that a revised and polished version be reviewed and republished.

Thank you for your attention to this matter.

 

 

 

[DelageBaza, Christina (DOA)]

Hi Caterina,

Here is a link to the metadata which has the specs, model, and contractor/subcontractor name.   I do not know the contact for the person/team who processed the data but I would first talk with Aimee/Chuck for their feedback.  https://www.arcgis.com/sharing/rest/content/items/58445d31693b4f80846c6f0e837c2859/info/metadata/metadata.xml?format=default&output=html

 

Christina Delage Baza, GISP
Data Analyst I
RIDOA - Division of Statewide Planning
235 Promenade Street

2nd Floor, Suite 230

Providence, RI 02908

 

401-222-6481

https://planning.ri.gov/planning-areas/data-center

 

 

 

From: Massidda, Caterina (EMA) <Caterina...@ema.ri.gov>
Sent: Wednesday, November 5, 2025 1:44 PM
To: DelageBaza, Christina (DOA) <christina....@doa.ri.gov>; Aimee Mandeville <aman...@uri.edu>; BCBSMA_labashc <lab...@uri.edu>; RIGIS <rigis...@uri.edu>; HoxsieQuinn, Garth (DOA) <Garth.Ho...@doa.ri.gov>
Cc: Washburn, John (EMA) <John.W...@ema.ri.gov>; Guthlein, Thomas (EMA) <Thomas....@ema.ri.gov>

Subject: RE: 2D Building footprint not accurate

 

Hi Christina,

Thank you for including them. I had submitted a couple of hundred feedback items back in July, so it was quite easy to notice that many of the same errors are still present.

Unfortunately, it seems that the AI model may not have been updated or retrained using the feedback provided by the Data Analysis group. Could you please share the contact information for the person or team who was in charge of this process?

Best regards,
Caterina

 

 

[Aimee Mandeville]

Good afternoon Caterina & Christina,

The building layer was generated using automated macros, many of the errors that were noted using the validation application were all related to the limitation of this automated solution.  Here is a summary of the communication with Fugro regarding the validation errors.

_______________________________

The RI stakeholder comments collected through the interactive ArcGIS web mapping app were provided to Fugro summarizing the errors of omission and commission.  Fugro indicated “further improvement of the classification quality really needs that extra human assistance – manual work”.  Fugro offered to review and assess the level of effort required to improve those specific areas (at additional cost).  We reported that all feedback (~348 points; 349 polygons) identified repeating patterns associated with solar panels, shipping containers, boats classified as buildings and occurrences where buildings may have been omitted.  

At the end of August, Fugro responded that most of the (app feedback) calls are related to the limitation of the automated solution defined through the project scope (Task Order) and can only be corrected with manual editing (not a requirement in Task Order). 5 instances can be corrected by applying automated macros.

The decision was made that it was best that Fugro not apply additional automated macros to fix the 5 instances so as not to potentially result in unintended damage to other areas. 

________

Best,

Aimee

[DelageBaza, Christina (DOA)]

Aimee,

Thank you for your feedback and communication summary with Fugro.  Maybe at the RIGIS EC meeting we can mention, if anyone tackles removing these incorrect features for their project needs, to share updates up to RIGIS to improve this layer for all?  

Christina Delage Baza, GISP
Data Analyst I
RIDOA - Division of Statewide Planning
235 Promenade Street 

2nd Floor, Suite 230

Providence, RI 02908

 

401-222-6481

https://planning.ri.gov/planning-areas/data-center

 

---

From: Aimee Mandeville <aman...@uri.edu>
Sent: Wednesday, November 5, 2025 2:47:53 PM

To: DelageBaza, Christina (DOA) <christina....@doa.ri.gov>

Cc: BCBSMA_labashc <lab...@uri.edu>; Massidda, Caterina (EMA) <Caterina...@ema.ri.gov>; RIGIS <rigis...@uri.edu>; HoxsieQuinn, Garth (DOA) <Garth.Ho...@doa.ri.gov>; Washburn, John (EMA) <John.W...@ema.ri.gov>; Guthlein, Thomas (EMA) <Thomas....@ema.ri.gov>
Subject: Re: 2D Building footprint not accurate