FG_CD_CSO_QandA_20260713v2.xlsx
XLSX spreadsheet 85 KB Posted
- Solicitation number
- HM047626SC001
About this file
This is a Questions and Answers document for the Foundation GEOINT (FG) Change Detection (CSO) Commercial Solutions Opening issued by the National Geospatial-Intelligence Agency (NGA) under the Department of Defense. The document addresses 50 vendor questions and provides NGA clarifications on the requirements and expectations for Phase 1 of the program. Key technical requirements include delivering change detection capabilities that identify Foundation GEOINT-relevant changes (mapping, charting, navigation, infrastructure, transportation, land cover, and disaster impacts) while filtering out non-relevant changes (seasonality, time of day, pattern of life, order of battle) across large geographic areas simulating a global business model on an annual or nearly annual basis. NGA requires solutions to detect changes as small as 100 square meters, deliver quarterly global change notifications over 3-month periods, and provide output as vector polygons in open standard GeoJSON format with metadata including date, source, location, areal extent, and confidence scores. The solution must leverage computer vision and AI/ML technology with transparent QA/QC process metrics achieving high industry standards for accuracy, though NGA has not specified acceptable false-positive or false-negative thresholds.
Critical operational parameters clarified include: offerors must source commercial or open-source imagery and computing power independently with no Government-Furnished Information (GFI) provided; the two-day latency requirement refers to ready access to data for analysis, not delivery timelines; the solution should function as a stand-alone data delivery compatible with standard government GIS formats and environments; and there is no limit to the total Area of Interest (AOI) or number of AOIs that will be run. Offerors must submit unclassified Solution Briefs using the provided template without deviations, include Team and Past Performance information, and propose multiple pricing options with lower-priced solutions favored. NGA has not determined whether any award will result from this CSO or established a timeline for subsequent phases; responses were due by July 21, 2026.
View the file
Other files
| File | Type | Posted |
|---|---|---|
| FG_CD_CSO_QandA_20260715.xlsx | XLSX spreadsheet | |
| FG_CD_CSO_QandA20260713.xlsx | XLSX spreadsheet | |
| FG_CD_CSO_QandA 792026.xlsx | XLSX spreadsheet | |
| Attch2_FG_CD_CSO_TOI_20260707.pdf | ||
| FG_CD_CSO_QandA_20260707.xlsx | XLSX spreadsheet | |
| Attch1_FG_CD_CSO_20260623v2.pdf | ||
| Attch1_FG_CD_CSO_20260623v2.pdf | ||
| Attch5_FG_CD_CSO_QandA.xlsx | XLSX spreadsheet | |
| Attch4_FG_CD_CSO_Clause_List_20260623.pdf | ||
| Attch3_FG_CD_CSO_SD_Template_20260623.pdf | ||
| Attch2_FG_CD_CSO_TOI_20260623.pdf |
Show all 11
On GovTribe
Work with this file
- Download the original file
- Contacts named in this file
- Similar government files
- Ask GovTribe AI about this file
Text version
CD-CSO
| CD-CSO-Vendor Questions | ||
| Question Numbers | Comment/Question | NGA Response |
| 1 | To support solution sizing and ROM pricing, can NGA provide the anticipated annual demand for on-demand change analysis requests? Specifically, what is the expected annual area of analysis (square kilometers), typical AOI size, and approximate number of requests, independent of the quarterly global change notification requirement? | As mentioned in the FG CD CSO TOI, “major challenge in this process is determining how to most efficiently understand if change has occurred to prompt an update to our products” and “Automated recurrent global change detection would inform future FDT data”. The purpose of this CSO will be to help NGA to gage foundation GEOINT relevant change and omit non-relevant on a global scale. It is yet to be determined the periodicity of accomplishing the FG group level global change detection requirements. Annual frequency and size of requests will be highly dependent on DOW and IC priorities, world events, and existing currency of data over these areas. The ideal solution should provide a cost-efficient process for identifying GEOINT relevant change over large regions of the earth, when combined, provide global coverage change detection on an annual or nearly annual basis. |
| 2 | The Topic of Interest primarily describes the delivery of a commercial change detection capability and associated data products. Beyond product delivery, does NGA anticipate requiring implementation or operational support services (e.g., system integration, deployment support, training, on-site support, or other professional services) as part of any resulting award? |
| Clarification on the expected balance between commercial product delivery and implementation services will help offerors propose an appropriately scoped solution and accurately address the Viability evaluation factor. | The primary purpose of the FG CD CSO will be to show detection capabilities of foundation GEOINT relevant change and omit non-relevant on a global scale. Agile change detection solutions that are compatible with standard government GIS formats and environments would typically offer the most viable solutions. While there is no requirement for implementation or operational support services, in the FG_CD_CSO_20260623 document Section A, there mentioned “Solutions may include existing technologies or procedures that are not currently in use that would enhance or streamline mission capabilities.” | ||
| 3 | The bulleted list under "Problem Statement" includes, "Detection of change over areas as small as 100 sq. meters." Does 100 sq. meters here refer to the AOI extent or the area of change detected? | The 100 sq meters refers to minimum areas of change. | |
| 4 | Is there a limit to the number of AOIs that will be run on the program? | There is no limit to the number of AOI’s. The purpose of this CSO is for identifying GEOINT relevant change over large regions of the earth, that could be accomplished through a multitude of smaller AOI’s or fewer larger AOI’s. | |
| 5 | Is there a limit to the total AOI area that will be run on the program? | While there is no limit to the total AOI area, for the CSO it will be a large area or series of smaller areas that when combined could start to simulate a global business model. | |
| 6 | Is there a limit to the total AOI area that will be run on the program? | NGA response is the same as for Question 5. | |
| 7 | Is there a timeline for the next phases for this effort-- Pitch, Full Proposal and Award? | NGA has not made a determination if there will be any award resulting from this CSO, or a timeline. | |
| 8 | Does the Agency intend to offer any GFI to support vendor proposals? | NGA has no intention of offering any GFI in support of the CSO at this time. | |
| 9 | What ground cover increments in square kilometers are expected to satisfy the CSO requirements? 250k, 500k, 1M, 25M, 50M, 100M, Global? | The TOI specifies possible areas as small as 100 sq meters, otherwise, there is no limit to the total AOI area. The CSO will require a large area or series of smaller areas that when combined could start to simulate a global business model. | |
| 10 | To optimize the change-detection deliverable, can NGA describe how the GeoJSON output is intended to be consumed downstream — e.g., ingested into a production or tasking system, reviewed by analysts, or persisted in some way? | NGA at this time will perform analyst review of the GeoJSON output to determine identification of FG relevant change. | |
| 11 | Does NGA maintain established confidence or accuracy expectations associated with specific feature classes, such that a change-detection confidence score should align to existing per-feature benchmarks rather than a single method-derived value? | In the FG CD CSO, NGA did not specify a specific threshold concerning confidence, accuracy or acceptable false-positive and false-negative rates. However, in the TOI Problem Statement, it specifies the proposed solution should include “transparent QA/QC process metrics ensuring the highest confidence in accuracy of the data”. While there is no specified number, NGA expects offerors to leverage technology to achieve high industry standards in this area. | |
| 12 | What are the target acceptable false-positive and false-negative rates for the fully automated stage — i.e., how accurate does detection need to be before a case gets flagged for analyst validation instead of passing through automatically? | In the FG CD CSO, NGA did not specify a specific threshold concerning confidence, accuracy or acceptable false-positive and false-negative rates. However, in the TOI Problem Statement, it pecifies the proposed solution should include “transparent QA/QC process metrics ensuring the highest confidence in accuracy of the data”. While there is no specified number, NGA expects offerors to leverage technology to achieve high industry standards in this area. | |
| 13 | Does the pipeline need to natively ingest and normalize multi-sensor inputs (e.g., comparing a PlanetScope 3-meter visual image to a Sentinel-2 10-meter spectral tile), or will operational input assets always come from identical sensor constellations? What is the expected resolution range (min & max)? | The primary purpose of the FG CD CSO will be to show detection capabilities of foundation GEOINT relevant change and omit non-relevant on a global scale. While there is no specific requirement concerning sensor and resolution, desired solution attributes should include “Leverage and continuously improve use of computer vision AI/ML technology”. | |
| 14 | Is there an operational requirement to filter out changes based on specific target classes (e.g., ignoring heavy commercial vehicles parked on hardstanding while flagging permanent concrete extensions)? | A desired solution attribute of this CSO is to leverage collaborative commercial market solutions to identify broad area FG-relevant change, while removing nonrelevant. | |
| 15 | What hardware limitations should we consider, and what are the specs of the system you’ll be testing on? | For this CSO, NGA has no hardware or system specs at this time other than to identify broad area FG-relevant change, while removing nonrelevant in vector, open standard formats with metadata, such as polygons in GeoJSON format. | |
| 16 | What data should be used for this project — open-source/commercial imagery sourced by the offeror, or will imagery/data be provided by NGA? | NGA has no intention of offering any GFI in support of the CSO at this time, open source/commercial imagery sourced by the offeror should be used. | |
| 17 | When the TOI refers to "historical" data, is comparison against the single most recent clear image sufficient, or does the baseline need to be composited from several older images (useful for persistently cloudy regions)? | For the purpose of this CSO, the baseline will be 3 months, in most cases, going backwards for “comparison against the single most recent clear image” may be sufficient to identify broad area FG-relevant change, while removing nonrelevant. When the “single most recent clear image” does not suffice, NGA expects demonstrations of agility as specified in the TOI Solutions Attributes and adjust and go back further as needed. | |
| 18 | Who supplies the imagery? Will this be Gov’t sourced or commercially sourced as part of the vendor’s innovative approach? | NGA has no intention of offering any GFI in support of the CSO at this time, open source/commercial imagery sourced by the offeror should be used. | |
| 19 | Who supplies the computing power? Is there a preferred platform for computing e.g. AWS? | NGA has no intention of offering any GFI in support of the CSO at this time, computing power by the offeror should be used. | |
| 20 | What is the initial GEO graphic extent? Whole world or specific AOI/AOR? | The initial GEO graphic area or AOI, for the CSO will be a large area or series of smaller areas that when combined could start to simulate a global business model. | |
| 21 | What type of imagery? Multi-imagery EO, SAR, IR. Etc. | The primary purpose of the FG CD CSO will be to show detection capabilities of foundation GEOINT relevant change and omit non-relevant on a global scale. While there is no specific requirement concerning sensor and resolution, the FG_CD_CSO_20260623 document Section A, points out “obtain innovative solutions” and the TOI points out the “objective of this effort is to leverage collaborative commercial market solutions”. | |
| 22 | Can we reference past high-side performance? Would be ideal to highlight success of high-side app development and Machine Learning (ML) modeling and implementation/deployment. | The FG CD CSO Solution Document includes a Team & Past Performance page that could be leveraged as long as it is compliant with the FG_CD_CSO_20260623 document Section B stating, “electronically submit…an unclassified Solution Brief”. | |
| 23 | Is there a specific timeframe for the initial deliverable product? E.g. 90-days, Quarterly, etc. (does it vary from the periodicity?) | Outlined in the TOI Problem Statement, “Change notifications should be scalable to global coverage which will provide a deliverable of relevant global changes detected over a 3 months period of time delivered quarterly”. | |
| 24 | Is the preferred solution a stand-alone solution (e.g. desktop or web app) or something that will be integrated into existing systems (e.g. Docker container, etc.)? | The FG CD CSO output at this time will function as a stand-alone data delivery solution. | |
| 25 | Does NGA anticipate making multiple awards under this CSO to support complementary capabilities (e.g., data collection, change detection, analytics, exploitation), or is NGA seeking a single end-to-end commercial solution? | NGA has not made a determination if there will be any award resulting from this CSO, or a timeline. | |
| 26 | Are there metadata standards or schemas that NGA expects offerors to follow? | Outlined in the TOI Problem Statement,” Vector indicators of change in open standard formats, attributed with metadata such as date, source, location, areal extent, and confidence score in standard formats”. | |
| 27 | What does NGA's on-demand analysis requests look like - e.g., API? | The FG CD CSO output will be more of a stand-alone data delivery at this time. | |
| 28 | Does NGA envision this capability becoming part of the Foundation Digital Twin operational baseline, or will it remain a stand-alone commercial service? | NGA has not made a determination if there will be any award resulting from this CSO, or a timeline. | |
| 29 | NGA indicated that additional pricing options would be useful. Could NGA please elaborate on the specific types of pricing options it would like industry to provide? Examples may include resolution-based pricing, square kilometer tiering, differentiated product features, delivery cadence, or other variables NGA would like offerors to consider." | The FG CD CSO will consist of AOI’s that simulate a global business model and as outlined in the FG_CD_CSO_20260623 document Section D, under possibility of funding, “solutions with lower prices are favored”. | |
| 30 | The TOI requests "change analysis based on ready access (2 days or less) to global data" but does not define expected request volumes, AOI sizes, turnaround expectations, or whether this capability is intended as an analyst-driven service, self-service capability, or demonstration of data availability. Can NGA clarify the expected operational concept to support development of an appropriate technical approach and pricing model? | The FG CD CSO is intended to be a stand-alone data delivery, offering CD solutions compatible with standard government GIS formats and environments. The FG CD CSO is pursuing FG-relevant change on a global scale and while there is no limit to the total AOI, it will consist of large or series of smaller areas that when combined could simulate a global business model. This operational concept should employ computer vision, and AI/ML technology advancements that contribute to state of the art, cost-effective solutions that bring best value to the government. | |
| 31 | What factors will NGA use in determining whether a selected solution will result in a FAR-based fixed-price commercial contract versus an OTA? | ||
| Reference: Attch1_FG_CD_CSO_20260623v2 Section B.A.2 pg.3 | NGA has not made a determination if there will be any award resulting from this CSO, or a timeline. | ||
| 32 | Will NGA provide access to all required current and historical imagery? | NGA has no intention of offering any GFI in support of the CSO at this time. | |
| 33 | How many concurrent users should we assume for viewing, reviewing, filtering, commenting on, and exporting AOI change-detection results? | That specific information has not yet been determined, for the purpose of this FG CD CSO, NGA expects delivery of the FG-relevant change data in formats as specified in the TOI. | |
| 34 | The government has provided a template for the offeror to utilize in responding to this requirement. Is it permissible to combine slides (e.g., have one slide that contains both introduction and opportunity responses) and still conform to the CSO. | To ensure conformance with the CSO, it is not permissible to deviate from the Solution Document template. | |
| 35 | The TOI distinguishes between relevant change for mapping, charting, navigation, infrastructure, transportation, land cover, and disaster impacts, and non-relevant change such as seasonality, time of day, pattern of life, and order of battle. Can NGA provide additional priority weighting among relevant change categories, or should offerors propose their own relevance taxonomy and confidence scoring approach? | While advancements in more detailed FG relevant detection capabilities are welcome, for the purpose of this CSO, there is not an expectation for the offerors to capture details from the thousands of features in the libraries of the FG domains. | |
| 36 | The TOI requests vector indicators of change in open standard formats with metadata including date, source, location, areal extent, and confidence score. Does NGA have a preferred schema, geometry type, attribution model, coordinate reference system, or compatible standard for the Phase 1 deliverable, such as GeoJSON, OGC API Features, shapefile/geopackage, STAC-linked metadata, or another FDT-aligned format? | For the purpose of this CSO, NGA has no specific preferred data format other than indicators of change in open standard formats, attributed with metadata such as date, source, location, areal extent, and confidence score in standard formats. | |
| 37 | For Phase 1, may offerors rely on commercially available EO/SAR imagery, public/open imagery, customer-licensed imagery, or a combination of these sources? Should the proposed solution include imagery acquisition as part of the fixed-price solution, or may it demonstrate the commercial processing capability using offeror-provided representative data? | The imagery solutions from question 37 would be sufficient and should be included as part of the fixed-price solution as NGA has no intention of offering any GFI in support of the CSO at this time. | |
| 38 | The TOI requests ready access within two days to global historical and current data. Should this be interpreted as access to the underlying imagery/catalog needed to perform analysis, delivery of change results within two days after AOI/time-window tasking, or both? | For the purpose of this CSO, the two days refers to ready access to the data to run the analysis on, not the analysis and delivery timeline. | |
| 39 | Does NGA have preferred QA/QC metrics for evaluating broad-area change detection, such as precision/recall, false positives per square kilometer, minimum mapping unit, geolocation error, areal overlap/IoU, confidence calibration, analyst review burden, or timeliness? | In the FG CD CSO, NGA did not specify a specific threshold concerning confidence, accuracy or acceptable false-positive and false-negative rates. However, in the TOI Problem Statement, it specifies the proposed solution should include “transparent QA/QC process metrics ensuring the highest confidence in accuracy of the data”. While there is no specified number, NGA expects offerors to leverage technology to achieve high industry standards in this area. | |
| 40 | For the quarterly global-change deliverable, does NGA expect comprehensive wall-to-wall global scanning, prioritized monitoring of foundation-interest geographies, or a scalable architecture and sampling-based demonstration during Phase 1? | The FG CD CSO will consist of AOI’s that-combined simulate a global business model. It is expected the FG-relevant change detected within the AOI’s will be identified in a manner such as polygons in GeoJSON format. | |
| 41 | Does NGA intend for the resulting capability to ingest and update an external, NGA-hosted persistent feature registry, or should offerors include a decoupled, standalone feature-state management repository within their solution boundary? If external, what standard interface (e.g., OGC API - Features) must the architecture support? | The FG CD CSO is intended to be a stand-alone data delivery, offering CD solutions compatible with standard government GIS formats and environments. It will consist of AOI’s that-combined simulate a global business model. It is expected the FG-relevant change detected within the AOI’s will be identified in a manner such as open standard polygons in GeoJSON format. | |
| 42 | To support automated feature reconciliation, will NGA provide open schema definitions (such as NSG Application Schema/NAS) to selected offerors post-award, or must the solution's internal semantic-context engine map outputs to arbitrary, vendor-defined ontologies? | NGA has not made a determination if there will be any award resulting from this CSO, or a timeline. If awarded suggestions and advancements in technology and efficiency could be possible. | |
| 43 | Is the terminal output of the proposed capability defined as a machine-to-machine vector change product routed directly to an existing NGA review system, or must the offeror deliver a dedicated User Interface (UI) to support manual analyst adjudication and state-transition approval workflows? | The FG CD CSO is intended to be a stand-alone data delivery. It is expected the FG-relevant change detected within the AOI’s will be identified in a manner such as open standard polygons in GeoJSON format. | |
| 44 | For the annual global change detection scope referenced in Q1, will NGA establish explicit geographic exclusion zones (e.g., open ocean, unpopulated regions), or must the system evaluate surface characteristics uniformly across all global landmass shapes? | For the purpose of this CSO, the AOI will consist of large and small areas that when combined would simulate a global business model. It will likely contain geographic exclusion areas and no open ocean to increase cost -effectiveness. | |
| 45 | Post-award, will operational commercial imagery ingestion be facilitated via centralized NGA enterprise access points (e.g., GIMS, Cedar, or Global EGD APIs), or are offerors expected to independently arrange, fund, and maintain direct commercial constellation harvesting pipelines? | NGA has no intention of offering any GFI in support of the CSO at this time. | |
| 46 | Does the "accessible within two days" latency threshold define the maximum allowable duration from initial satellite sensor acquisition to final change product delivery, or does it govern internal processing time once the data becomes available inside the processing environment? | For the purpose of this CSO, the two days refers to ready access to the data to run the analysis on, not the analysis and delivery timeline. | |
| 47 | Should the solution architecture be engineered with plug-and-play microservices to ingest and orchestrate external third-party CV/AI model detections, or should the platform be optimized exclusively around independent, solution-owned analytical methods? | The FG CD CSO is intended to be a stand-alone data delivery. It is expected the FG-relevant change detected within the AOI’s will be identified in a manner such as open standard polygons in GeoJSON format. The requirement is for “continual process improvements through computer vision, and AI/ML technology” and for the purpose of this CSO, “our interest is not in the identification of features, only identification of change”. | |
| 48 | Does NGA require a unified, mandatory temporal change-detection algorithm across all surface environments, or does the agency favor an adaptive approach where different temporal comparison methods (e.g., CCDC vs. BFAST) are dynamically assigned based on regional land-cover types? | As indicated in the TOI, for the purpose of this CSO, NGA requires detection of FG relevant change “includes changes applicable to mapping, charting, and navigation such as land use/land cover; air, land, and sea transportation and navigation features and networks; urbanization indicators; and land/infrastructure change due to natural disaster’ omission of non-relevant “includes changes related to time of day, seasonality, pattern of life, and order of battle”. The requirement is for “continual process improvements through computer vision, and AI/ML technology” and for the purpose of this CSO, “our interest is not in the identification of features, only identification of change”. | |
| 49 | Are there baseline performance metrics (e.g., minimum Recall or a maximum False Alarm Rate per square kilometer) required for technical acceptability, or are offerors evaluated on their capability to dynamically output confidence scores that let NGA adjust thresholds operationally? | In the FG CD CSO, NGA did not specify a specific threshold concerning confidence, accuracy or acceptable false-positive and false-negative rates. However, in the TOI Problem Statement, it specifies the proposed solution should include “transparent QA/QC process metrics ensuring the highest confidence in accuracy of the data”. While there is no specified number, NGA expects offerors to leverage technology to achieve high industry standards in this area. | |
| 50 | To support appropriate security engineering and ROM sizing, what are the target deployment parameters for the operational capability? Specifically: (a) What cloud fabric and impact level (e.g., C2E Top Secret) is designated, and (b) Will personnel require facility access or specific clearances during the initial development phase? | For the purpose of this CSO, it is intended to be a stand-alone data delivery and the FG-relevant change detected within the AOI’s will be identified in a manner such as open standard polygons in GeoJSON format. NGA expects responses to be compliant with the FG_CD_CSO_20260623 document Section B stating, “electronically submit…an unclassified Solution Brief”. |
UNCLASSIFIED
UNCLASSIFIED
File details come from the government source that posted it. Updated .