The Space Domain Awareness Challenge


The Aerospace Corporation and TechCrunch are joining forces to host a pitch competition to find the strongest startups using AI/Machine Learning to work with satellite data streams. Space Domain Awareness arises from the output of all the sensor data.

Space operators are totally reliant on sensors in space and on the ground to get a complete picture of what’s happening in Earth’s orbit and beyond. Space operators must task sensors correctly with the complex job of collecting data. The challenge is then parsing, packaging and disseminating the output of a significant number of data streams to the correct data consumer across various teams at the correct time.

We call this scenario in Space Domain Awareness the “Media Streaming Problem” because, just like popular streaming providers, the data has to be collected, analyzed and delivered to the correct audience at the right time. 

The Central Question:

How can AI/ML tool sets be used to parse sensor data, address the Media Streaming Problem and increase our understanding of satellite sensor generated data?

Challenge Data Sets:

Applicants should be able to work with the following data sets. Experience in working with complex data streams, whether in space or not, is needed.

Telemetry Data: All satellites have telemetry data. Telemetry data does not come down in a repeatable, cyclical way. It is tasked for downlink depending on what’s in the system – sometimes it’s which way the spacecraft is pointing, sometimes it’s satellite bus temperature. The features in the data change over time and do so in a non-recurring fashion. Data scientists call this non-stationary, time-series data. To use an email queue example, the email coming into you depends on what messages you send out, with some random messages coming in. 

Mission Data: Sensors capture and transmit data back to people to figure out what’s happening in space. The space-based data comes from Radio Frequency (GPS), imagery (Earth Observation/Infrared, Synthetic Aperture Radar, hyperspectral), Overhead Infrared (OPIR is different than EO/IR – looking for heat and brightness vs. image). Ground-produced data includes: optical telescopes, radio telescopes, and radar data.

Comms Links Data: This allows mission data to get to where it needs to go. Typically satellite comm is included in the telemetry process.

Open to early-stage companies who are in “Idea Stage” up to and including “Series A Stage” companies who are also either:

  • Wholly-owned US company with a US-based workforce
  • Company based in a NATO member country (including Australia, Japan, New Zealand, Norway) AND authorized to participate in NATO related business opportunities


For application submission questions, email