Podcast Summary: CyberWire Daily – AWS in Orbit: Automated Satellite Management

Episode Details

• Title: AWS in Orbit: Automated Satellite Management
• Series: T-Minus Space Daily
• Hosted by: Maria Varmazes
• Guests: Dax Garner (CTO, Cognitive Space) and Ed Miletian (Solutions Architect, AWS Aerospace and Satellite)
• Release Date: May 26, 2025

Introduction

In this episode of the AWS in Orbit series from the 40th Space Symposium, host Maria Varmazes engages in an insightful discussion with Dax Garner, CTO of Cognitive Space, and Ed Miletian, Solutions Architect at AWS Aerospace and Satellite. The conversation centers around the advancements and challenges in automated satellite operations, leveraging artificial intelligence (AI) and machine learning (ML) to optimize satellite management at scale.

Meet the Guests

• Dax Garner introduces himself as the CTO of Cognitive Space with a robust background in aerospace engineering, including experience at NASA Johnson Space Center and Firefly. He highlights his passion for integrating AI and ML into space infrastructure to facilitate human spaceflight in the future.

  "Cognitive Space's mission is to empower the use of space assets, particularly developing AI and ML algorithms for proliferated systems and mission management of those assets."
  — Dax Garner [03:06]

• Ed Miletian discusses his role at AWS Aerospace, bridging cloud solutions with aerospace needs. With experience spanning NASA and Space Force missions, Ed emphasizes his focus on mission management, planning, and leveraging cloud scalability to support complex space operations.

  "Deployment like this is really, really crucial for these critical missions that our governments have."
  — Ed Miletian [16:12]

Cognitive Space: Mission and Challenges

Dax elaborates on Cognitive Space’s mission to manage proliferated satellite systems—constellations comprising hundreds of satellites working cohesively for missions like Earth imaging or establishing global mesh networks. The primary challenge lies in solving combinatorial optimization problems efficiently.

• Combinatorial Optimization Example:

  Managing 100 satellites to capture optimal images of Earth involves deciding which satellite should photograph specific targets at any given time. Traditional algorithms struggle with the vast number of possibilities, often leading to inefficient heuristic approaches that sacrifice optimality for speed.

  "The sweet spot is training and designing ML models that can run at the speeds of heuristics... but we can gain back a lot of optimality."
  — Dax Garner [06:47]

• Dynamic Scheduling:

  Satellite operations are dynamic, with potential task additions or hardware failures necessitating rapid re-optimization of schedules. ML models enable swift adjustments, maintaining mission integrity without the delays inherent in traditional optimization methods.

  "If you're losing tasks or gaining new tasks... you have to re-optimize on the system, otherwise you're dropping collections."
  — Ed Miletian [07:07]

AI and ML Enhancing Satellite Operations

Dax explains how Cognitive Space utilizes AI and ML algorithms hosted on AWS infrastructure to handle large-scale optimization tasks:

• Efficiency Gains:

  Cognitive Space’s ML models achieve 90-95% of optimal solutions compared to traditional heuristics, which often operate at 50-60% optimality. This significant improvement translates to better mission performance, such as increased image collection and more effective communication link management.

  "We're really getting back to like 90, 95% of the optimum solutions... planning hundreds of satellites in minutes."
  — Dax Garner [12:04]

• Real-World Applications:

  • Space Development Agency (SDA): Optimization of link management for their mesh network constellation.
  • National Geospatial-Intelligence Agency (NGA) and National Reconnaissance Office (NRO): Enhancing geospatial request handling by integrating commercial satellite capacities.

  "We are helping them understand that capacity and make predictions about whether they can fulfill certain requests on the commercial side."
  — Dax Garner [08:07]

AWS's Role in Enabling Scalable Solutions

Ed highlights how AWS’s cloud infrastructure underpins Cognitive Space’s operations, providing the necessary scalability and security:

• Scalability:

  AWS services like EC2 and ECS enable Cognitive Space to scale their ML models seamlessly, accommodating the onboarding of new satellite constellations without the need for system overhauls.

  "The architecture... scales really well to train the models and also to execute the models when you're planning."
  — Ed Miletian [10:33]

• Security:

  Given the sensitive nature of clients such as SDA, NGA, and NRO, AWS ensures robust security protocols, including support for GovCloud and highly classified environments.

  "These services all can run in both our GovCloud regions as well as our most classified secret and top secret regions."
  — Ed Miletian [11:40]

• Cloud Mission Operations Center:

  Ed introduces the concept of a cloud-based Mission Operations Center, which integrates various mission subsystems, allowing for scalable and flexible mission planning and execution without the constraints of physical hardware management.

  "With the cloud Mission Operations center, you are automatically getting the best available technology under the hood."
  — Ed Miletian [16:12]

Impact and Future Outlook

Real-World Impact:

Dax shares benchmarking results where Cognitive Space’s ML algorithms outperform traditional heuristics, offering near-optimal solutions swiftly. This capability is crucial for handling complex, large-scale satellite operations efficiently.

"We're planning in minutes... whether that's establishing link schedules or planning sensor management and collecting pictures on the ground."
— Dax Garner [12:04]

Future Developments:

• Agentic Systems:

  The integration of ML models with agentic systems enables dynamic decision-making, allowing autonomous adjustment of objectives and constraints based on real-time operational conditions.

  "An agent will start to decide... this model might be the best way to resolve and replan."
  — Dax Garner [15:54]

• Expanding Constellations:

  With the growth of satellite constellations, Cognitive Space aims to support both government and commercial startups in managing increasingly complex space assets without the overhead of developing proprietary management systems.

  "AWS's Cloud Mission Operations center is already there ready to onboard their proliferated systems as they fly their first, second and then eventually become their full scale constellations."
  — Dax Garner [18:51]

• Space Economy Enablement:

  Facilitating the broader space economy by providing scalable, secure, and efficient management tools for satellite operations, Cognitive Space envisions a future where managing large satellite networks is seamless and integrated.

  "The key is to enable an entire space economy... we can help them mission management."
  — Dax Garner [18:51]

Conclusion

The episode underscores the pivotal role of AI and ML in transforming satellite management, making operations more efficient, scalable, and adaptable to dynamic conditions. AWS’s robust cloud infrastructure provides the foundation necessary for Cognitive Space to deliver cutting-edge solutions that meet the evolving needs of both government and commercial space sectors. As satellite constellations continue to grow, collaborations like that of AWS and Cognitive Space will be instrumental in driving the future of space operations.

Notable Quotes:

• "The sweet spot is training and designing ML models that can run at the speeds of heuristics."
  — Dax Garner [06:47]

• "Deployment like this is really, really crucial for these critical missions that our governments have."
  — Ed Miletian [16:12]

• "We're really getting back to like 90, 95% of the optimum solutions."
  — Dax Garner [12:04]

About the Podcast:

CyberWire Daily by N2K Networks delivers the latest cybersecurity news and analysis, featuring interviews with industry leaders from around the globe. Hosted by Maria Varmazes, each weekday episode delves into critical topics shaping the cybersecurity and aerospace industries.