Industry_Questions_1_-_DOIGFBO250002_2.pdf

Artificial Intelligence/Machine Learning (AI/ML) for Landsat Flight Operations

Clarification Questions and Responses

DOIGFBO250002

Q1. How much investment is expected from non-federal entities (aka industry)?

A1. The investment from the partner depends on the scope of cooperative work being envisioned and the current capabilities of the vendor (e.g. what resources are in place, what is new/additional, etc.). U.S. Geological Survey (USGS) is not looking for a specific level of investment, but a partner is expected to provide sufficient resources to enable a successful Cooperative Research and Development Agreement (CRADA) effort.

Q2. What is the rationale for treating it as a public-private partnership vs procuring services from industry?

A2. USGS is still in the "research" phase of considering AI/ML and plans to work with industry to cooperatively develop systems that may be applicable to satellite/mission operations.

Thus, USGS doesn’t have a formal set of requirements to evaluate a procurement activity against. As it is expected that there will be many of avenues for implementation, the USGS feels a CRADA is the best way to explore industry capabilities and concepts, potentially including some that may be too immature to support a funded (procurement) activity.

Q3. What are the key challenges or inefficiencies in current Landsat satellite flight operations that AI/ML applications are expected to address?

A3. This is provided in the announcement within the “Purpose and Objectives”:

• “The USGS is interested in AI/ML technologies applied to areas including, but not limited to, anomaly triage and telemetry trending and analysis, and welcomes information in other areas of flight operations to consider.”

• “This announcement is issued to help identify potential collaborators for a CRADA to assess if and how AI/ML techniques can enhance decision making and operational efficiency to streamline satellite operations and improve mission reliability. Expected outcomes include a better understanding of the efficiency gains in flight operations through use of AI/ML technology and the feasibility of implementing AI/ML solutions across different satellite systems.”

AI/ML for Landsat Flight Operations

DOIGFBO250002

Q4. What metrics will be used to evaluate the success or feasibility of AI/ML solutions in optimizing flight operations?

A4. The metrics for evaluation would be developed as part of the process in putting a CRADA into place. The purpose of this sources sought is to determine interest in putting such an agreement into place; the USGS and the partner would work jointly to define objectives and expectations.

Q5. What level of access will respondents have to USGS or NASA subject matter experts during the project?

A5. The USGS has engineering support to help describe and package telemetry data, as well as meet and answer any clarifying questions throughout the project. The level of access would depend on the scope of the cooperative work proposed. Note that the required level of support for government resources could be an evaluation criterion in the decision to execute a CRADA, as the USGS has a limited supply of resources to apply to the task.

Q6. How will the USGS assess the feasibility and potential impact of proposed AI/ML approaches during the planning phase?

A6. The USGS will review the proposed scope of work and make assessments based on criteria such as: resources needed, schedule, objectives and expected outcomes, area of investigation, feasibility of work proposed to achieve desired outcomes, experience of vendor in this area, etc. In general, the USGS is interested in techniques and capabilities to improve satellite operations by enhancing performance and optimizing costs in any area of flight operations.

Q7. Are there specific qualifications, certifications, or prior experiences that will strengthen a respondent's submission?

A7. Previous experience with AI/ML algorithms and methods with similar types and volume of data would be advantageous. Previous experience with satellite operations to include detailed analysis of telemetry data or other satellite operations applications would be viewed as a strength.

Q8. What is the timeline for the review of submissions and the potential establishment of a

CRADA?

A8. Review of responses will occur after the RFI response date through the end of the calendar year. We expect to start the process of putting agreements in early in calendar year 2025.

AI/ML for Landsat Flight Operations

DOIGFBO250002

Q9. How does this initiative align with broader USGS and NASA goals for satellite?

A9. The USGS is continually looking for techniques and capabilities to improve satellite operations and optimize costs, while providing higher quality performance and capabilities.