NMFS_-_STEREO_CAMERA_-_Request_for_Information_1.docx

NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION FISHERIES

REQUEST FOR INFORMATION

A Compact, Low-Cost, Wide Angle, Underwater, Stereo Imaging System This Request for Information (RFI) is issued for market research purposes in accordance with Federal Acquisition Regulations (FAR) Part 10. It does not constitute a solicitation, nor does it guarantee that a solicitation will be issued. Requests for a solicitation or proposals in response to this RFI will not be entertained, and this RFI does not obligate the government in any way.

The Department of Commerce (DOC), National Oceanic and Atmospheric Administration (NOAA) Fisheries, Pacific Islands Fisheries Science Center (PIFSC), and Western Acquisition Division (WAD) are seeking information on the development of a compact, low-cost, wide angle underwater stereo imaging system. This system is intended to be a key component of an automated, end-to-end optical data acquisition and analysis pipeline for fishery-independent surveys, supporting NOAA Fisheries’ mission to advance stock assessments, ensure sustainable fisheries, and contribute to the New Blue Economy.

BACKGROUND:

Underwater camera systems have become essential tools in fisheries research, offering a non-extractive, fishery-independent means of surveying fish populations in their natural habitats. Over the past decade, advancements in optical technology and machine learning have significantly expanded the utility and efficiency of these systems. They address many limitations of traditional survey methods, such as those dependent on extractive sampling or visual observation from crewed vessels.

As NOAA Fisheries moves towards platform-agnostic, uncrewed survey systems, the integration of advanced optical hardware with automated data processing pipelines is increasingly critical. Tools like VIAME, CoralNET, and many others illustrate the potential of machine learning to enhance data utility but require closer hardware-software integration, including capabilities for real-time processing (e.g., AI at the edge).

Recent challenges, such as reduced days at sea (DAS) aboard NOAA Fishery Survey Vessels (FSVs) and the need to collect data in otherwise inaccessible areas (e.g., wind lease sites), underscore the importance of transitioning to scalable, flexible optical systems. Compact, low-cost, and easily deployable systems will be pivotal in addressing these challenges while expanding data collection capabilities.

NOAA envisions a transformative imaging system that is small, lightweight, affordable, and capable of wide angle stereo imaging. This system should integrate seamlessly into an automated pipeline, from data acquisition to analysis, to support fishery-independent surveys. Utilizing commercially available, off-the-shelf components is encouraged to reduce costs and ensure long-term industry support for maintenance and scalability.

By embedding integrated hardware and software solutions into real-time onboard processing systems, NOAA aims to improve the efficiency, flexibility, and quality of its monitoring efforts. This approach will reduce reliance on NOAA FSVs, diversify sampling strategies, and extend monitoring capabilities to under-served regions, transforming the NOAA Fisheries monitoring enterprise.

CURRENT EXISTING TECHNOLOGY USE CASE EXAMPLES

The Pacific Islands Fisheries Science Center deploys the Modular Optical Underwater Survey System (MOUSS) to estimate the species-specific, size-structured abundance of commercially-important fish species in Hawaii and the Pacific Islands (Amin et al. 2017). The MOUSS is an autonomous stereo-video camera system designed for the in situ visual sampling of fish assemblages. This system is rated to 500 m and its low-light, stereo-video cameras enable identification, counting, and sizing of individuals when range to target is 0.5–10 m. The modular nature of MOUSS allows for the efficient and cost-effective use of various imaging sensors, power systems, and deployment platforms. The MOUSS has been used for surveys in Hawaii, the Gulf of Mexico, and Southern California. In Hawaiian waters, the system can effectively identify individuals to a depth of 250 m using only ambient light. The MOUSS has been in operational use for more than ten (10) years, and many of the components are past their end of life and are difficult to source or repair. Technology has markedly improved since development of MOUSS, hence, the need to develop the next generation system.

The Southeast Fisheries Science Center has been deploying 360º-stereo video cameras since 2016 as part of the Southeast Area Monitoring and Assessment Reef Fish Video (SEAMAP-RFV) survey that dates back to 1992. This survey has been further developed to the Gulf Fishery Independent Survey of Habitat and Ecosystem Resources (GFISHER) - a gulf-wide survey which provides relative abundance for population indices of 24 federally managed species of commercial and recreational fishing value (Switzer et al. 2022). Further experimentation with 360º systems across a variety of habitat types, ocean basins, and taxa has demonstrated their effectiveness to improve accuracy of fish counts as well as provide a more comprehensive understanding of the habitat at deployment sites (Kilfoil et al. 2017, Campbell et al. 2018, Kilfoil et al. 2021). The wide angle camera arrays provide means for monitoring species in untrawlable habitats on natural and artificial reefs within 10-200m of depth throughout the northern Gulf of Mexico. Cameras record monochrome videos at 10 frames per second for higher contrast in low, ambient light conditions. This specific camera design has been in use by NOAA and other partner organizations since 2016, and has been the standard for the GFISHER survey since then.

MINIMUM TECHNICAL SPECIFICATIONS for IMPROVED SYSTEM:

The overall objective is to develop the smallest, lightest, easy-to-use, and cost-efficient system possible to collect species-specific, size-structured data on target fish populations.

Required (minimum viable product)

1. Fully autonomous and self-contained: all data acquisition, data storage, and power functions residing within the unit.

2. Robust ability to withstand and maintain stereo-calibration during transport, deployment, and recovery in rough sea conditions from a variety of vessel platforms. This may include significant shocks to the system, including being drug across rocky seafloor habitats.

3. Easily deployable by a team of two persons aboard vessels as small as 19ft, by hand.

3.1.Highly desirable: Easily hand-deployable by a single individual in any vessel of opportunity.

4. Easily operable by minimally trained, non-technical staff (e.g. commercial fishers) with little to no contact with developers, technical staff, or advanced training materials.

4.1.Highly desirable: as easy to operate as standard off-the-shelf camera technology, requiring no training or training manuals for successful basic usage. All necessary manuals or training material required for advanced/full usage provided in clear written manuals and/or training videos.

5. Positive system function verification and status prior to and immediately following deployment (e.g. Stats on the deployment (how many images captured, max depth, dwell time, maybe even the ability to view a sampling of images from the deployment).

6. Fully serviceable by entities other than the developer, including NOAA hardware/software engineering staff.

7. Dry weight: ≤ 25 lb.

7.1. Highly desirable: As light as possible.

8. Working Depth rating: 500 m

8.1. Highly desirable: 750+ m

9. System standby time: ≥ 24 hours

10. Video recording time: ≥ 5 hours

11. Low- and high-ambient-light performance (e.g. low: 3.4x10-4 lux, f/0.8)

a. Strong sensitivity long wavelengths (red-orange and red wavelengths; 590 - 740 nm range)

b. Highly desirable: Monochrome imager

c. Highly desirable: Ability to control artificial lighting

12. User-adjustable recording parameters using lossless file compression within industry standard file containers

12.1. Ability to record video

12.1.1. Video specifications:

12.1.1.1. single view resolution: ≥ 1920 X 1080 pixels

12.1.1.2. frame rate: ≥ 10 Hz

12.2. Highly desirable: Record still images in addition to video

12.2.1. Still image specifications

12.2.1.1. single image resolution: ≥ 3.2 Megapixels

12.2.1.2. frame rate: ≥ 6 fps

13. Internal recording of ambient depth (pressure), temperature.

13.1. Highly Desirable: Ability to interface with diverse external modules (e.g. data sonde, hydrophone, transmissivity and turbidity sensors, current speed/direction). Including extra bulkheads to incorporate additional sensors, lights or cameras.

14. Lenses or lens systems in the form of two ~90-135º x 60º field of view (FoV), back to back facing orientation that provide an unobstructed 180-270º total FoV with stereo-coverage allowing for measurement of target individuals within the 180-270º FoV.

14.1. Highly desirable: 360º-stereo coverage allowing target measurement orthogonally or across the full 360º FoV, with ability to add or remove cameras as required

15. Camera control software that has the capacity to add additional cameras as desired by researchers needing more than 270º coverage or additional external cameras pointed in directions of interest (e.g. downward facing, or satellite pairings).

16. Stereo camera baseline (inter-camera spacing), low lens distortion, and stereo-camera synchronization sufficient to allow measurement of target organisms with a precision of +/- 2cm in the Z-axis at a range of 4m (at optimal target placement).

16.1. Highly desirable: Indexed, variable, user-adjustable, camera baseline separation allowing for up to +/- 0.5cm target measurement precision in the Z-axis at a range of 7.5m across the field of view.

17. Users must be able to download image data from the camera system without needing to open the pressure housing.

17.1. Highly desirable: Wireless and hardwired external network connection

18. Software system allows user configuration for recording parameters e.g. duration, file naming

18.1 Manual end of recording shut off with maintained file integrity Highly Desirable

1. Supports near-real time, in-situ, image processing using “AI-on-the-edge” or similar capabilities. Examples include real-time object detection, enumeration, and classification, using externally trained models.

2. All hardware and software should be expected to have a 10-year service and support life and should be readily serviceable by NOAA staff or external 3rd party vendors.

POTENTIAL DELIVERABLE

Deliverable 1: Provide meeting/logistics and administrative consulting services, including:

a. Set up design consolation meetings and hardware exchanges (prototype delivery).

b. Share progress on hardware design and manufacture through virtual meetings every two months.

c. Conduct equipment testing for robustness, pressure rating, and durability

d. Set up a demonstration meeting that can be attended by multiple NOAA program representatives to show available equipment.

e. Set up prototype consolation meetings to discuss NOAA evaluation and refinements leading to the final unit.

Deliverable 2: The contractor shall provide a working prototype of the Compact, Low-Cost, Wide Angle, Underwater, Stereo Imaging System at readiness level 7.

Deliverable 3: A final report including full technical details, design drawings, CAD files, system and software specifications, and any and all source code necessary to allow the government the option to use, replicate, and/or modify the hardware and software developed for operation of the described camera system.

SPECIAL REQUIREMENTS:

Qualifications The Contractor must demonstrate that they have the requisite experience in the following:

· Underwater camera hardware design and fabrication

· Underwater battery systems and lighting

· Managing equipment logistics, including shipping, storage, maintenance

· Availability for communications with clients during field deployments

· Working with marine fisheries information programs or other science-based programs

· Basic machine learning and computer vision systems, include edge AI.

The contractor shall provide interdisciplinary expertise related to program and project management, business processes, best practices and procedures.

1st article testing 1st Article Testing and approval may be appropriate when-

(a) The contractor has not previously furnished the product to the Government;

(b) The contractor previously furnished the product to the Government, but-

(1) There have been subsequent changes in processes or specifications;

(2) Production has been discontinued for an extended period of time; or

(3) The product acquired under a previous contract developed a problem during its life;

(c) The product is described by a performance specification; or

(d) It is essential to have an approved first article to serve as a manufacturing standard.

Vendors are encouraged to submit responses/proposed solutions. WAD anticipates a single award with approximate values between $100k-$365k.

The NAICS code for this requirement is anticipated to be 334511, with a size standard of 1350 employees.

Response Requirements Responses to this notice shall be limited to ten pages:

Organization Information

1. The organization name and address, and a point of contact name, telephone number and e-mail address. The organization point of contact shall have the knowledge and authority to clarify responses with Government representatives.

2. The organization’s UEI number, CAGE code, and organization/company structure (corporation, LLC, partnership, joint venture, etc.).

3. Written confirmation of the organization’s registration in SAM (www.sam.gov), if applicable. Offerors need not be registered to be considered as a source, but must be active to participate in any subsequent solicitations, if issued.

4. The size classification of the organization under the appropriate NAICS code, and if the organization is classified as a small business, HUBZone small business, Service Disabled Veteran Owned small business, Woman-owned small business and/or 8(a) certified small business.

5. General statement providing the organization’s relevant experience and/or capability to provide the described services, with an emphasis on the following information:

1. Anticipated timeline for tailoring software for specific business processes/pipelines

2. Software description(s)

3. Rough Order of Magnitude for initial development

4. Experience with DOC systems, within NOAA or other Bureaus Capability statements and other responses shall be submitted to Nathan Rowsell at nathan.rowsell@noaa.gov no later than 2:00 PM Pacific Time on Friday, December 20th 2024. Please do not call regarding this RFI. Any questions regarding this notice must be submitted in writing via e-mail to the above email address not later than 2:00 PM Pacific Time on Friday, December 7th 2024. An amendment may be posted if necessary, so firms are encouraged to “follow” this announcement for future actions.

This request is not to be construed as a commitment on the part of the Government to award a contract, nor does the Government intend to pay for any information submitted as a result of this request. No entitlement to payment of direct or indirect costs or charges by the Government will arise as a result of the submission of contractor's information. The Government shall not be liable for or suffer any consequential damages for proprietary information not properly identified. Proprietary information will be safeguarded in accordance with the applicable Government regulations.

REFERENCES

● Amin R, Richards BL, Misa WFXE, Taylor JC, Miller DR, Rollo A, Demarke C, Singh H, Young GC, Childress J, et al. 2017. The Modular Optical Underwater Survey System. Sensors. 17:1–14. doi:10.3390/s17102309.

● Campbell, M.D., Salisbury, J., Caillouet, R., Driggers, W.B. and Kilfoil, J., 2018. Camera field-of-view and fish abundance estimation: A comparison of individual-based model output and empirical data. Journal of Experimental Marine Biology and Ecology, 501, pp.46-53.

● Kilfoil, J.P., Wirsing, A.J., Campbell, M.D., Kiszka, J.J., Gastrich, K.R., Heithaus, M.R., Zhang, Y. and Bond, M.E., 2017. Baited Remote Underwater Video surveys undercount sharks at high densities: insights from full-spherical camera technologies. Marine Ecology Progress Series, 585, pp.113-121.

● Kilfoil, J.P., Campbell, M.D., Heithaus, M.R. and Zhang, Y., 2021. The influence of shark behavior and environmental conditions on baited remote underwater video survey results. Ecological Modelling, 447, p.109507.

● Switzer TS, Keenan SF, Thompson KA, Shea CP, Knapp AR, Campbell MD, Noble B, Gardner C, Christman MC. 2022. Integrating Assemblage Structure and Habitat Mapping Data into the Design of a Multispecies Reef Fish Survey. Marine and Coastal Fisheries: Dynamics, Management, and Ecosystem Science. 2023:1-24. DOI: 10.1002/mcf2.10245