Long-range visual observation system

Updated 1 March 2021: FAQ section update with new Q&A

Executive Summary

The Department of Defense seeks a long-range, integrated multi-spectrum camera system, computer vision, and edge compute capability for detection, identification, observation, and warning. The solution must be capable of tracking and identifying objects (e.g. aircraft and vehicles) in real time from distances of 30 nautical miles or more. Intended use cases will require day, night, and all-weather functionality. Proposed solutions should include automated analytic functions -- such as image enhancement, object recognition, bandwidth reduction, or other functionality -- ideally through on-device or edge compute analysis. These functionalities should be governed by a common user interface and aid operators’ ability to quickly detect and classify objects of interest. The proposed solution, if successful, should be capable of integrating with third party systems such as visual warning systems, additional optics, software analytics, etc. Finally, the solution should transmit data securely using open-source, non-proprietary formats.

Given the range of capabilities sought by the Department, teaming between hardware companies (e.g. specializing in edge-compute or multi spectral camera systems) and software companies (e.g. specializing in computer vision or multispectral analysis) is highly encouraged.

Product Requirements

The DoD seeks to prototype a readily available solution that includes the following capabilities: 

• Provide multi-spectrum real-time observation at distances of 0 to at least 30 nautical miles
  
  • Spectrums: electro-optical (EO) and infrared (IR) required, others optional
  • System must include the ability to report the range and direction to designated objects of interest
  • Example range and resolution for object detection: system should be able to detect an airliner with 36 pixels at 30 nautical miles
  • Cameras should be capable of tracking subsonic moving objects

• Use on-device or edge compute processing, machine learning, and analytics to improve operators’ ability to detect and monitor objects of interest. Some examples may include, but are not limited to:
  
  • Image enhancement
    
    • Filter or penetrate high aerosol atmospheric conditions, such as fog or haze
    • Improve performance in dark and low-light conditions
    • Image stabilization and correction to mitigate turbulence from wind or weather
  • Object recognition, identification, and alerting
    
    • Automatically identify and flag objects of interest
    • Classify object type
    • Calculate range and azimuth to objects of interest
    • Calculate geographic coordinates for objects of interest
  • Automated object of interest tracking 
  • Reduce bandwidth requirements
    
    • Image compression

• Successful prototypes will be integrated into existing command and control networks and data management applications and must be able to:
  
  • Securely transmit data and receive instructions via Transport Layer Security wired and, optionally, wireless connection.
  • Securely transmit all data in common, non-proprietary data formats.
  • Zoom, focus and interrogate objects of interest under remote operator control.

• Finally, we invite vendors to propose a solution mounted in an all-weather housing and control unit capable of: 
  
  • Operating 24/7/365 outdoors in adverse conditions, such as high wind, salt fog, snow, high temperatures, humidity, lightning storms, dust, solar radiation, and fungal growth conditions.
  • 360 degrees of continuous panning movement, and tilt from -5 degrees to 40 degrees.
  • Remote control.
  • Mounting and supporting additional third party systems (e.g. a visual warning system or temperature control systems) that can potentially weigh up to 150 pounds. 

Teaming is Encouraged

Preference will be given to submissions that present a comprehensive and compelling solution to the entire AOI. Proposals should identify whether the submitter will employ subcontractors and, if so, which companies for which capabilities. 

However, companies that have an innovative approach to specific aspects of the AOI are encouraged to submit proposals. Such companies, if selected, may be asked to arrange a prime / subcontractor relationship with other submitters to ensure the delivery of an integrated solution. 

Successful solutions will demonstrate consistency with the DoD AI Ethical Principle.

Frequently Asked Questions:

Q: How will this system be deployed?

A: Initial deployment of this system will be on a stationary, ground-based platform.

Q: What does the system being able to detect an airliner with 36 pixels at 30 nautical miles mean?

A: The system should be able to register 36 horizontal pixels on the target (HPoT)

Q: What does multi-spectral in the solicitation mean? 

A: The system must have electro-optical (EO) and infrared (IR) sensing capability, other spectrums are optional.