Amazon Web Services (AWS) is strategically positioning its cloud infrastructure to leverage the potential of generative AI in transforming space and other industries. With over 60% of AWS' space and aerospace customers already using AI, the company predicts a significant increase in the adoption of generative AI in the coming years. AWS has restructured internally to focus on generative AI, creating a dedicated “generative AI for space” team and a laboratory for customers to experiment with new applications. The company sees three main areas for generative AI in space: geospatial analytics, spacecraft design, and constellation management.

Emiliano Kargieman, CEO of Satellogic, has written a really interesting post on how the synergy of Earth observation and AI is set to revolutionize our understanding of the planet. Something I’ve touched on a few times.

He has posted a Twitter thread on the topic as well.

Rocket Lab and True Anomaly were just selected for an extremely cool tactically responsive mission by Space Systems Command. Per Rocket Lab’s press release the two teams will demonstrate the capability to develop space vehicles with rendezvous and proximity operation (RPO) functionality, as well as establish corresponding command and control centers for their operation in a mission called VICTUS HAZE:

Once the spacecraft build is complete, Rocket Lab will be entered into a Hot Standby Phase awaiting further direction. Once the exercise begins, Rocket Lab will be given notice to launch the spacecraft into a target orbit. After reaching orbit, the spacecraft will be rapidly commissioned and readied for operations. Rocket Lab will configure a Pioneer class spacecraft bus to meet the unique requirements of the VICTUS HAZE mission and launch the spacecraft on Electron from either Launch Complex 1 in Mahia, NZ or Launch Complex 2 in Wallops, VA. Once on orbit, the spacecraft will conduct a variety of dynamic space operations to demonstrate SDA characterization capabilities with True Anomaly’s spacecraft, the Jackal autonomous orbital vehicle.

Rocket Lab’s Pioneer spacecraft was previously used to support the Varda mission.

True Anomaly’s AI-enabled Jackal spacecraft is really interesting as well:

Ars Technica has a comprehensive piece up with a more detailed profile of the exercise:

“When another nation puts an asset up into space and we don’t quite know what that asset is, we don’t know what its intent is, we don’t know what its capabilities are, we need the ability to go up there and figure out what this thing is,” said Gen. Michael Guetlein, the Space Force’s vice chief of space operations.

This is what the Space Force wants to demonstrate with Victus Haze. For this mission, True Anomaly’s spacecraft will launch first, posing as a satellite from a potential adversary, like China or Russia. Rocket Lab will have a satellite on standby to go up and inspect True Anomaly’s spacecraft and will launch it when the Space Force gives the launch order.

“Pretty sporty,” said Even Rogers, co-founder and CEO of True Anomaly.

Then, if all goes according to plan, the two spacecraft will switch roles, with True Anomaly’s Jackal satellite actively maneuvering around Rocket Lab’s satellite. According to the Space Force, True Anomaly and Rocket Lab will deliver their spacecraft no later than the fall of 2025.

There’s another interesting aspect to this mission mentioned in a Payload article:

Planners expect to increase complexity to the point where one vehicle will actively avoid being characterized. The goal isn’t just to demonstrate the technology, but also develop the tactics and procedures for these kinds of missions.

The demonstration is a collaboration between the Defense Innovation Unit, Space Safari acquisition program, and SpaceWERX.

The proof of concept will launch in 2026:

…the cubesat aims to set a new standard for computing power on a small platform, said SAIC. By enabling AI and machine learning “at the edge” in space, the satellite can process data with minimum delay.

With more compute power in orbit, there is less need for transporting data back to the ground for processing and relay to users,” said a SAIC spokesman. “The latency of decision making is vastly reduced.”

Pushing as much of the identification, classification, and even prioritization to the edge as we can is going to be a game changer on the intelligence front especially as we reach a state where there is persistent surveillance and change detection.

Payload has a really interesting piece on the state of earth observation (EO) and how the industry is shifting its focus towards profitability, government funding, and addressing niche issues, as commercial demand lags expectations.

Key points:

• Current EO market breakdown: ~50% US government, 25% allied governments, 25% commercial
• Major players like Planet, Spire, BlackSky, and Maxar are prioritizing profitability after years of growth
• Early movers have an advantage with significant initial investments creating barriers to entry
• Startups need to differentiate through niche capabilities like hyperspectral imaging, climate monitoring, or edge computing
• AI and data analytics are seen as potential growth drivers for the industry
• SpaceX’s $1.8B NRO contract highlights the government’s dissatisfaction with the state of EO and their desire for cost-effective EO solutions

Of course the macro-economic environment that many of these companies have come to market in hasn’t helped. As far as I’m concerned the focus on profitability and selectivity in new venture funding is a welcome side effect of these new economic realities. The survivors from the current crop will grow stronger and the industry should finally be able to avoid the boom bust cycle that has plagued previous attempts to cement new space as a persistent and critical part of the economy. The commercial markets will grow - just never as quickly as the pitch decks predict.

Mission name: Beginning Of The Swarm
Launch Vehicle: Electron
Launch Site: Launch Complex 1
NZT Launch Window: 10:00, April 24, 2024
UTC Launch Window: 22:00, April 23, 2024
ET Launch Window: 18:00, April 23, 2024
PT Launch Window: 15:00, April 23, 2024
Mission Overview: Rocket Lab’s ‘Beginning Of The Swarm’ mission is set to launch from New Zealand during a 14-day window starting on April 24th. The Electron rocket will carry two satellites for separate customers: NEONSAT-1, an Earth observation satellite developed by SaTReC and KAIST, and NASA’s Advanced Composite Solar Sail System (ACS3).

NEONSAT-1, the primary payload, is designed to monitor natural disasters along the Korean Peninsula using high-resolution optical imagery and artificial intelligence. The secondary payload, NASA’s ACS3, will demonstrate new materials and deployable structures for solar sail propulsion systems, which use sunlight for propulsion instead of conventional rocket propellant. The solar sail will measure approximately 30 feet per side when unfurled, and the flight data will be used to design future larger-scale composite solar sail systems for various space missions.
Live Stream: rocketlabusa.com/livestream

For Additional Updates: Follow Rocket Lab’s official Facebook and Twitter accounts.

Voyager Space and Palantir have teamed up for another AI+Space partnership:

Working with a company like Palantir, well known for its artificial intelligence and machine learning technology, would facilitate processing in space, allowing for a more concise transmission of data to Earth.

“At Voyager, we have a lot of heritage in space station work, exploration work, the design and development,” Kuta said. “But one area that it definitely made sense to partner with a company like Palantir was on the software side.”

Kuta said the work with Palantir will be focused primarily on national security that could include other allied nations and partner organizations such as NATO.

“Voyager and Palantir have already done joint proposals to potential customers,” Kuta said.

As posted earlier, I expect this trend to only increase as companies move to lock in some of these existing data pipelines and position themselves to build new ones.

I wrote a post a few days ago about the importance of AI and New Space for each other and said that the realization may sink in with the public and wall street types more broadly once someone like OpenAI makes a big move in the space. Well, it turns out that someone is NVIDIA:

The field of weather prediction is changing rapidly due to advances in AI. Spire’s multipurpose constellation of satellites, utilizing RO technology, offers unmatched global coverage of precise vertical profiles of temperature, pressure, and humidity across the globe, including under-observed areas and remote regions. By combining this dataset and Spire’s proprietary DA analyses with NVIDIA’s advanced AI platforms and APIs, this collaboration aims to usher in a new era of accuracy in weather forecasting.

“As our world grapples with dynamic shifts in climate, anticipating and mitigating the impact of extreme-weather events poses escalating challenges for companies, individuals, and communities,” said Michael Eilts, general manager of weather and climate at Spire. “Aligning Spire’s proprietary data and unmatched global weather coverage with NVIDIA’s cutting-edge technology and expertise positions us to markedly elevate the accuracy of weather prediction. This collaboration will help ensure our customers are not just informed but empowered to proactively address the evolving climate landscape.”

Spire Global, like many New Space companies, possesses a crucial asset sought after by AI solution providers—massive amounts of high-quality, useful, and hard-to-source data. While this RO data has numerous applications, Spire Global has invested heavily in leveraging it for weather and climate analytics as well as forecasting. The company is already making significant strides in this domain, but partnerships with AI leaders like NVIDIA could unlock markets and benefits far beyond those initially envisioned by New Space data generators. Expect to see more groundbreaking announcements along these lines in the near future.

I asked GPT-4 to generate a list of 30 business concepts that leverage LEO observation data and AI:

• High-Resolution Advertising Analytics: Monitors and analyzes foot traffic in commercial areas using high-resolution satellite imagery, offering insights to advertisers for targeted billboard placements.
• Offshore Energy Monitoring: Provides surveillance and maintenance planning for offshore energy installations like wind farms and oil rigs, enhancing efficiency and safety.
• Smart Logistics and Supply Chain Management: Uses satellite data to optimize shipping routes by monitoring weather conditions and maritime traffic, reducing fuel consumption and improving delivery times.
• Deforestation Alert System: Offers real-time alerts on illegal deforestation activities, catering to governments, NGOs, and carbon offset companies for immediate action.
• Land Dispute Resolution Services: Utilizes historical satellite imagery to resolve land disputes by providing undisputable evidence of land use and boundaries over time.
• Luxury Crop Monitoring: Specializes in monitoring the health and status of luxury crops like vineyards and truffle farms, providing precision agriculture services to high-value agriculture sectors.
• Mining Exploration and Management: Supports mining operations by identifying potential sites, monitoring environmental compliance, and managing resources more efficiently with satellite imagery.
• Construction Progress Tracking: Offers construction firms and investors the ability to remotely monitor project progress and compliance with plans, reducing the need for onsite visits.
• Polar Route Navigation Services: Provides up-to-date ice and weather data for ships navigating through polar routes, optimizing travel time and ensuring safety in increasingly accessible Arctic and Antarctic passages.
• Virtual Real Estate Tours: Creates detailed, satellite-based virtual tours of real estate properties, especially in remote or undeveloped areas, enhancing marketing efforts for real estate agencies.
• Infrastructure Health Monitoring: Uses high-resolution imagery to assess the condition of critical infrastructure like bridges, roads, and dams for maintenance planning.
• Agricultural Insurance Verification: Verifies insurance claims by monitoring crop health and damage after natural events, ensuring accurate claim processing.
• Space Debris Tracking and Management: Identifies and tracks space debris to mitigate collision risks with operational satellites and spacecraft, supporting safer space operations.
• Precision Agriculture: Offers detailed crop health monitoring and soil moisture levels, enabling farmers to optimize irrigation, reduce waste, and improve crop yields.
• Climate Change Monitoring: Utilizes satellite imagery to track changes in ice caps, deforestation rates, and desertification, providing data for climate research and environmental policy-making.
• Urban Expansion Analysis: Tracks urban sprawl and infrastructure development, aiding city planning and sustainable growth initiatives.
• Disaster Response and Management: Provides real-time data on natural disasters like floods, wildfires, and hurricanes, improving emergency response and recovery efforts.
• Ocean and Marine Life Monitoring: Studies changes in marine ecosystems, tracks illegal fishing activities, and monitors ocean pollution levels.
• Air Quality and Pollution Tracking: Uses observation data to map pollution sources and air quality, supporting environmental health initiatives.
• Forest Management and Protection: Detects illegal logging activities, monitors forest health, and supports reforestation efforts by analyzing vegetation cover over time.
• Wildlife Migration Tracking: Observes and analyzes wildlife migration patterns and habitat changes, aiding in species conservation efforts.
• Glacier Monitoring Services: Tracks glacier retreat and melting rates, offering vital data for water resource management and climate science.
• Real Estate and Land Use Analysis: Provides insights into land use changes, real estate development patterns, and valuation trends for investors and developers.
• Energy Infrastructure Monitoring: Monitors the construction and operation of renewable energy farms (solar, wind) and traditional energy facilities for efficiency and environmental impact assessments.
• Archaeological Site Discovery: Helps in identifying potential archaeological sites and changes in historical landscapes through remote sensing technologies.
• Insurance Risk Assessment: Uses data on environmental and urban changes to better assess risks related to natural disasters and climate change, impacting insurance premiums and policies.
• Iceberg Tracking for Shipping: Offers real-time data on iceberg locations and movements, enhancing navigational safety for maritime shipping routes.
• Water Resource Management: Monitors reservoir levels, river flows, and overall water resources to support sustainable management and allocation.
• Snow Cover and Avalanche Prediction: Analyzes snowpack levels and stability to predict avalanche risks, supporting ski resort management and public safety initiatives.
• Coastal Erosion Monitoring: Tracks changes in coastal lines and erosion rates, aiding in coastal management and protection strategies.

BlackSky is doing some interesting work in the AI space:

BlackSky, a supplier of satellite imagery and space-based intelligence, won a $2 million U.S. defense contract to provide data to train AI models.

The contract was awarded by the defense contractor Axient on behalf of the Air Force Research Laboratory. Axient in September 2023 won a contract from AFRL worth up to $25 million for space experiments.

Axient will use BlackSky’s satellite imagery and data analytics platform to support studies and technology demonstrations focused on tracking moving objects from space.

As someone who works on AI solutions professionally, I’m a little surprised that there isn’t more discussion about the immense value and potential of space data and AI. The value is pretty well understood with constellation builders and in the intelligence sector but less so with the general public, the broader commercial sector, and on Wall Street. Companies like Blacksky, Spire Global, and other LEO operators are generating vast libraries of data that, when made available to AI for training or analysis, will drive countless impactful applications. There is currently a lot of focus on real-time or fresh data but for many applications the value of these libraries grows significantly as they age and scale.

Low-cost sensor platforms and launch have fueled the rise of New Space but the convergence of AI and these capabilities is the key to unlocking hugely profitable markets that will cement space as a significant and permanent pillar of the economy - one that is visible and top of mind for the general consumer. I expect this realization to sink in within the next couple of years, or sooner, if someone like OpenAI makes a high profile move in this space (which I am also expecting) much like they did recently with robotics.