Perceptive Space raises $2.8M to predict space weather using AI

Perceptive Space, a Canadian aerospace startup, has emerged from stealth mode with $2.8 million in funding to predict space weather.

The Toronto company it will use the pre-seed funding in the oversubscribed round to develop a revolutionary space weather platform aimed at providing enhanced predictions and decision intelligence.

Space weather, which includes dynamic conditions in the space environment such as solar flares and geomagnetic storms, can severely disrupt satellite operations, communication systems, and terrestrial power grids. With satellites growing in importance for things like Starlink internet access, it’s clear there’s a need for knowing what space weather will be like.

Currently, space weather information is provided primarily by government agencies like the National Oceanic and Atmospheric Administration (NOAA). However, these forecasts often lack the necessary accuracy, lead times, and decision intelligence needed for modern operations, Perceptive Space said.

AI-first approach

What distinguishes Perceptive Space is its AI-first approach and that allows them to augment traditional space weather forecasting techniques and improve the accuracy and provide near real updates. The company leverages AI and sensor fusion and analyzes decades of space weather and operational data and has demonstrated predictions that are up to 10 times more accurate than existing forecasts at the bench scale, said Padmashri Suresh, CEO of Perceptive Space, in an email to VentureBeat.

Unlike traditional forecasting techniques, using AI-first approach enables lightweight prediction models that can support on-edge integrations for autonomous or automated systems.

“Our cutting-edge technology integrates data from multiple sources, utilizing AI algorithms to predict space weather and its potential impact on satellite operations, communications, and other critical infrastructure,” said Suresh. “By providing highly accurate forecasts and actionable insights, Perceptive empowers organizations to proactively manage and mitigate the effects of space weather, ensuring the resilience of their operations.”

Suresh founded Perceptive in 2022, drawing on her extensive experience in space weather modeling, Smallsat missions, space policy, and building and scaling AI products in the tech industry.

The early team — with five employees — includes leading experts in space weather, machine learning, and aerospace engineering, with backgrounds from prestigious institutions such as NASA, Los Alamos National Labs, and MIT, and experience developing AI-enabled products for high-profile clients including DARPA, Google, Meta, and AWS.

The recent funding will support Perceptive’s efforts to scale its technology. The round was backed by a group of investors specializing in AI and space technologies, including Panache Ventures, Metaplanet, 7Percent Ventures, Mythos Ventures and AIN Ventures.

Perceptive is targeting the space and defense markets space and defense markets in the United States and allied countries and has already secured early commitments from several satellite operators and launch providers, including a pilot, demonstrating strong market demand for their innovative solutions. The company is working on releasing their first product in 2025.

Space weather background

Space weather refers to the environmental conditions in space, primarily influenced by the sun. During space weather events, such as solar flares and geomagnetic storms, the space environment experiences extremely high levels of radiation and plasma.

These conditions can drag satellites out of their orbits, cause satellite failures, result in GPS signal outages, and disrupt spacebased operations. To mitigate these impacts, the space industry has to adjust
satellite orbits, reschedule launches, and temporarily shut down non-essential satellite systems.

In an email to VentureBeat, Suresh said she and other team members have been working on the research for over 10 years. Suresh did her doctoral research on the topic.

The problem today: relying solely on government

Today, the industry has to rely solely on agencies like NOAA for space weather predictions to take these mitigation actions. These predictions lack accuracy, reliability, and sufficient insights to help take suitable actions. Unlike terrestrial weather, where you have real-time and hyper-local weather predictions to help with aircraft operations, spacecraft operations are flying relatively blind to weather in space today.

Due to this, the industry has to to be reactive to space weather and it is underprepared to take the right action at the right time. This means the companies either don’t act early enough or shut down systems for durations longer than needed, resulting in service disruptions and financial losses that can easily run up to hundreds of millions of dollars.

Recent examples of space weather disruptions include SpaceX and Intelsat losing satellites, Capella Space and ESA satellites deorbiting prematurely, and reports of Iridium, Sen, and Planet pausing their operations to mitigate the impact of space weather.

The inspiration

Back in the mid-2010s, Suresh was in Washington, D.C., working on space policy at the Space Studies Board of the National Academies of Sciences. There were two things that she had a front-row seat to. That include the promise of a future led by a commercial space industry, which meant thousands of new satellites and new use cases for space in the near future for “a new space age.” And she also saw competing priorities and budgetary constraints that plague agencies like NOAA and NASA make them fundamentally slower and unable to match the pace of commercial space.

“Both led to the realization that the [doctoral] research I was working on — improving space weather using ML would become very important in the coming years as we enetred a “new space age,” and government agencies won’t be able to meet the space weather demands of this new commercial-led space industry,” Suresh said.

Suresh added, “However, I also knew we needed to hit a critical mass of scale for this commercial space weather company. The loss of 40 Starlink satellites in 2022 acted as the signal for me that we had hit this critical mass for me to start Perceptive Space.The fact that space weather could take out 40 satellites at once highlighted both the scale of the industry operations and the impact of space weather on these operations.”

As for AI, she said, “AI helps us extract more signals from decades of solar and satellite operations data than the current approach of using stand-alone physics or statistical models, and helps us predict things earlier and with better accuracy compared to these existing models.”

The future of the problem

As the space industry scales up to a post-Starship world of daily launches and 100,000 satellites crowding the orbits around Earth, insufficient space weather intel will seriously hinder our ability to operate.

This scale-up risk comes not just from the lack of predictions capable of supporting the increase in scale but also from the lack of capabilities to support emerging activities in space, such as space habitats and in-space manufacturing. For instance, there currently are no space weather models that can support the safety of sustained human presence on Mars or the moon.

Perceptive Space said if we continue to rely solely on government agencies like NOAA, which are plagued by budgetary constraints and lengthy development times, the gap between space weather capabilities and industry needs will continue to widen, threatening the future of humanity’s space ambitions.

The solution

Perceptive Space is bridging the gap in space weather predictions with its AI-powered platform. This platform leverages its AI-based space weather models and experience bringing AI products from concept to market.

The team comprises seasoned space weather scientists, machine learning scientists and aerospace engineers with research backgrounds from NASA, Los Alamos National Lab, MIT, and the University of Waterloo, and experience in building AI-enabled products used by enterprises such as DARPA, Google, Meta, and AWS.

The AI-based space weather predictions have outperformed traditional models by more than ten times in accuracy, speed, and reliability at the bench scale. With the support of investors such as Panache Ventures, Metaplanet, Mythos, 7percent Ventures, and AIN Ventures, the company is scaling this technology to support humanity’s space endeavors, from Very Low Earth Orbit (VLEO) to the Moon, Mars, and beyond.

The platform equips space and defense sectors with the tools needed to anticipate and mitigate the impacts of space weather on their operations. This enhanced intelligence enables better decision-making regarding whether to power through or shut down under various space weather conditions. By supporting both satellites and human space exploration missions, the company will ensure that space weather no longer hinders financial or intellectual advancements in space.

As humanity continues to reach for the stars, the company believes we must protect investments and infrastructure from the Sun’s fury. Without better space weather predictions, the bright future of the space industry may remain just out of reach.