Brief IA
›

Constellation Space Claims Up to 36% More Data

AI Tools·Tom Levy·

Constellation Space Claims Up to 36% More Data

Constellation Space Claims Up to 36% More Data
Key Takeaways
1Constellation Space announces a satellite link optimization platform based on machine learning
2The startup claims up to 36% more data transferred per pass and over 10 clients, including AWS
3More than 20 academic institutions are using the platform, with an annual contract value of $2.25 million
Why it matters : Automating data transfer is becoming crucial with the increasing number of satellites in low Earth orbit.

The Seattle-based startup is deploying a satellite link optimization platform that leverages machine learning. It claims throughput gains per pass and announces around ten clients, including AWS, as well as an ACV of $2.25 million. The founding team includes profiles from NASA, SpaceX, Blue Origin, and Capital One.

Claimed Gains and Initial Contracts Signed

Constellation Space asserts that its clients can increase the amount of data transferred during a pass by up to 36%. According to Kamran Majid, an early version of the platform attracted more than 10 clients, including Amazon Web Services. He also notes that over 20 academic institutions are already using the tool. The executive estimates the company's annual contract value at $2.25 million for this year.

Capacity Predictions and Machine Learning Orchestration

The platform utilizes machine learning to enhance data transfer and simplify the management of satellite fleets and terrestrial networks. It aggregates forecasts regarding situational awareness, weather conditions, signal-to-noise ratios, signal interference, and bandwidth demand to assess the volume of gigabytes that can be transmitted during each pass. These estimates help operators adjust settings and transmission time slots. The model is pre-trained on a global network of sensors and then adapts to the specific hardware of each fleet and ground stations. Kamran Majid specifies that the team builds the models for weather, attenuation, interference, and demand in-house, synthesizing them to assist operators in anticipating ground contact plans and optimizing their fleets.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

A Small Team from NASA, SpaceX, Blue Origin, and Capital One

Founded last year and based in Seattle, the company is led by CEO Kamran Majid, who was an engineering intern at NASA and a software engineer on Starlink at SpaceX. CTO Raaid Kabir is a veteran of Blue Origin, while Omeed Tehrani and Laith Altarabishi previously worked as software engineers at Capital One. Constellation Space has released a video showcasing the platform in operational conditions. Kamran Majid indicates that the team currently consists of six people.

Launch, Target Market, and Choice of Seattle Location

The company is deploying its platform for low Earth orbit constellations, a segment where tracking thousands of satellites and transfer windows is becoming increasingly complex. Seattle is presented as a region where more than half of the world's satellites are built. Kamran Majid recalls that, having gone through Y Combinator, the team spent a few months in San Francisco and felt pressure to stay there, but he believes Seattle is becoming a space hub, with companies moving in or opening their headquarters there. He also highlights the city's technological reputation and the team's attachment to the location. Reflecting on his experience, he mentions witnessing the degradation of manual processes at Starlink with the rise of constellations and explains that he created Constellation to help small teams anticipate link failures. He describes his transition from SpaceX to the startup as "really interesting" and outlines a mindset guided by first principles and the elimination of irrelevant requirements.

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.