Lidar_RFI_qs_Final.docx

Lidar RFI Questions and Answers

1. Is the USFS only looking at already existing COTS solutions?

The Forest Service is interested in COTS solutions that meet the described requirements. If respondent has a cost effective solution to meet requirements that is not a COTS, the Forest Service is interested in those solutions as well.

1. Will the USFS consider a custom designed and build implementation to meet their required specifications?

Yes, custom designed solutions would be considered. The USFS encourages all responses that meet required specifications.

1. Can the USDA Forest Service (USFS) expand on who their end-users are and how they currently use LIDAR data?

The Forest Service has a wide variety of end-users who use lidar data for research, analysis, and decision-making support. Lidar-related publications by research and development scientists in the Forest Service can be found on Treesearch: https://www.fs.usda.gov/treesearch/search?keywords=lidar&authorlname=&yearfrom=&yearto=&station=&series=&volume=

1. Does the USFS anticipate their enterprise system to be where end-users across all regions can conduct geospatial analysis or do they anticipate maintaining a decentralized, regional system?

To comply with the FHTARA mandate, the Forest Service is primarily looking for centralized cloud solutions.

1. Why do users need to download LIDAR data in bulk (e.g., what is the end-use of the data) and what type of file formats are they downloading data in?

Although anticipation is that a system can eventually be set up for full scale analysis of all sorts, the Forest Service recognizes that some of the analysis that scientists and others in the agency may want to do is outside of scope of the business requirement. Ability to view data and also be able to download for additional processing is a critical requirement.

1. Are there data science or analytics users at USFS who want to programmatically interact with LIDAR data? If so, what libraries and languages do they use?

Please see response to question 5.

1. What applications does USFS use to analyze LIDAR data?

The primary software used to analyze Lidar includes: ArcGIS, ERDAS, eCognition, and an internal FS produced application called FUSION.

1. How many different types of LIDAR data does the USFS currently have?

Lidar point cloud formats are .las, .laz, or .zlas (ArcGIS) in las version 1.2, 1.3, or 1.4. Raster derivatives are .tif or .img format. FS lidar is primarily airborne.

1. What granularity of LIDAR data does USFS currently have?

8 pulse/meter2 full point clouds and 50cm-1m raster derivative products.

1. How is the USFS currently dealing with processing different granularities of LIDAR data?

Most processing is done using Lidar Fusion, eCognition, Erdas Imagine, or ArcMap.

1. How does the USFS currently approach change detection in regards to LIDAR data processing?

Using differencing techniques. Currently the USFS is working on baseline lidar coverage and there is very little 2nd acquisition lidar for monitoring and change detection.

1. What other data does the USFS use in conjunction with LIDAR data for their workflows and analyses? For instance, is LIDAR data overlaid on other data sources?

Primarily digital airborne imagery, e.g. NAIP, and high-medium resolution satellite imagery, e.g. Worldview-3 and Landsat. Data may be overlaid with inherent ArcGIS image layers or Google Earth imagery.

1. In regards to accessing data, what are the bandwidth expectations for client uploads and downloads?

The FS hopes to learn what possibilities exist through responses to this RFI.

1. How is USFS expecting clients to connect to the central system in order to upload files?

Web interface API would be the ideal setup for access. If the respondent has other means could achieve access, those descriptions would be considered also.

1. What file formats need to be supported for data downloads?

Lidar point cloud formats are .las, .laz, or .zlas (ArcGIS) in las version 1.2, 1.3, or 1.4. Raster derivatives are .tif or .img format.

1. Does USFS have Service Level Agreements (SLAs) in place around downloads/uploads and processing time (e.g., how frequently are data uploads happening, how long should uploads or downloads take, etc)?

No, the USFS is not aware of any of these that are in place.

1. Does the USFS have a specific requirement or capability in mind around hardware or cloud infrastructure for this solution?

The Forest Service does not have a specific requirement for hardware or cloud infrastructure for the solution.

1. What file formats do USFS end-users need LIDAR data in?

lidar is .las, .laz, or .zlas. Raster derivatives are .tif or .img.

1. In regards to the creation of a “nationwide, best-available area," what heuristics does the USFS use to compute a best-available area (e.g., highest resolution, most recent, etc)?

This should be addressed in RFI responses.

1. What security model does USFS anticipate or prefer for the enterprise system (e.g., differentiating read/write/own users, creating Role Based Access Controls, creating Attribute Based Access Controls, etc)?

Ideally, this would be Role Based Access with Active Directory Authentication, where feasible. Other options would be considered.

1. Is there a centralized cross-agency set of groups/permissions for who can access what already in place at USFS?

No, this does not currently exist.

1. Which existing software programs are used to analyze LIDAR data?

The primary software used to analyze Lidar includes: ArcGIS, ERDAS, eCognition, and an internal FS produced application called FUSION.

1. Are there specific backup formats USFS requires for data?

No.

1. Does USFS have an estimation of the intended duration, or period of performance, associated with this RFI? Additionally, does USFS have an estimate of their budget for the opportunity associated with this RFI?

This request is for informational purposes only. Based on responses to this RFI, the Forest Service will review information described and will then, if appropriate budget and approvals can be obtained, will move forward with an RFQ. Certainty and dates of an RFQ are not known at this time.

1. Has GTAC applied the NIST Risk Management Framework to determine the information classification of data sets to be stored in the cloud data management platform? If so, what is the minimum FedRAMP classification required for the cloud Lidar data management system?

No, but the FS anticipates that data will be classified differently depending on its sensitivity.

1. Has GTAC determined a roadmap for when the cloud data management FedRAMP authorization must be achieved?

No, this has not been determined.

1. Is GTAC prepared to sponsor an Agency Authorization To Operate (ATO) for the cloud data management platform if the needed solution requires it?

Yes, if GTAC moves forward with an RFQ, an ATO will be required and CIO/GTAC will work to achieve that requirement.

1. In Section 4.d for dissemination, will all users be required to authenticate? If so, what is the acceptable authentication specification for Federal Agency users internal and external users to Forest Service? What is the acceptable authentication method for non-Federal Agency users?

The FS hopes to learn about potential options from responses to this RFI.

1. Is Forest Service willing to see a demonstration of suggested platforms to better evaluate what exists on the market today?

Yes, based on responses to this RFI, the Forest Service would be willing to see suggested platforms available.

1. Page 2 states a response page limit of 30 pages, whereas page 5 states there is no page limit for the response. Please clarify.

A. The announcement has been corrected to state that there is a 30 page limit.

1. Is there any particular cadence for data ingestion. For example, daily, X times per day, etc.

The FS hopes to learn about potential options from responses to this RFI.

1. For large batch ingestion (multiple TB), is it assumed this type of ingest will be “scheduled” or is it assumed to be simply “on demand”?

Ideally the Forest Service would like to have large batch ingests completed on demand. If this is not possible or respondent has other recommendations, the Forest Service is interested in that information.

1. What type of connectivity is to be assumed for large data ingest/downloads? Internet? Dedicated circuits? Other?

Forest Service is interested in what respondent recommends for large data ingests/downloads.

1. Is there more information that can be shared about the current data management approach in terms of the architecture, data processing and analytics capabilities? This is in reference to the statement “The USFS is seeking information on a system or systems that will replace most or all of the current data management approach.”

The current data management approach is ad hoc. The FS is interested in an enterprise-level solution for lidar data management.

1. Are there any specific performance problems with the current setup in terms of memory, computing and analytic processing needs?

The RFI is looking for responses that can meet business requirements. Descriptions of current limitations are not relevant.

1. Is there a need to combine the LiDAR point cloud and raster-derivative data with data from other sources? If yes, is there a present list of data sources with which data merging, feature engineering and data cleanup needs to be performed? Is there a wish-list of other sources that will be good-to-have as part of the data engineering?

Same answer as question 12.

1. Is there a present list and/or wish-list of analytic questions that the proposed solution’s visualization and predictive analytics component should be capable of addressing?

No. However, to illustrate the types of questions the FS is asking, Lidar-related publications by research and development scientists in the Forest Service can be found on Treesearch: https://www.fs.usda.gov/treesearch/search?keywords=lidar&authorlname=&yearfrom=&yearto=&station=&series=&volume=

1. “The USFS currently has more than 207 TB of lidar point cloud and raster-derivative data.”

37. What is the database technology currently used to store this data?

37. What is the expected data growth rate?

37. Of the 207TB, what proportion of data is analyzed on a regular basis and the remaining is maintained for archival purposes and accessed only for specific need with low frequency?

Some of the Lidar raster derivative products are served as Image Services using ArcGIS for Server platform tools on infrastructure currently housed in Salt Lake City. A portion of the data are also stored and accessible through the Forest Service Virtual Data Center in Kansas City. Growth rate is expected to be at least 10% per year, if not more. The proportion of data analyzed on a regular basis vs maintained for archival purposes is not known at this time.

1. Does Forest Service intend (now or in the near future) to use a public cloud for a long-term archive site?

Yes

1. Is encryption of the data going to be a requirement? De-duplication?

No

1. How granular of meta-data tagging does FS require?

The FS hopes to learn about potential options for metadata tagging through responses to this RFI.

1. What type of reporting does FS want to see from this system?

Please see Section 4d in the RFI.

1. What applications are in use today, or planned to be, that would analyze the data?

Same answer as Question 7.

1. How is data security and access managed today?

Please see response to question 34.