We’ve learned that Interstellar Objects (ISOs) are not strangers to our Solar System. Many have visited, and many more will in the future. The Vera Rubin Observatory is expected to find hundreds each year. Scientists are keen to learn more about them, and a swarm of spacecraft on standby might be the way to do it.
On a basic level, an ISO is simply an object unbound to any star. The two we know of are ‘Oumuamua which was detected in 2017, and Comet Borisov, detected in 2019. ISOs typically have very high velocities, follow hyperbolic trajectories that show they don’t orbit the Sun, and have unique compositions that set them apart from Solar System bodies. ‘Oumuamua, for example, could be a hydrogen iceberg, though this is just one possibility.
Scientists are eager to examine these objects closer and understand their compositions and origins. Unfortunately, their high velocities make them elusive, and we can only glimpse them with ground-based telescopes. What’s needed is a way to visit one. The best way to do that is to have a spacecraft waiting to catch up with one as it passes through the inner Solar System.
Or even better, a whole swarm of spacecraft that don’t require explicit instructions to rendezvous with an ISO.
Hiroyasu Tsukamoto is with the Department of Aerospace Engineering in the Grainger College of Engineering at the University of Illinois Urbana-Champaign. He and his colleagues developed Neural Rendezvous, a deep learning-driven guidance and control framework that can autonomously guide spacecraft to ISOs. Their work is in a paper titled “Neural-Rendezvous: Provably Robust Guidance and Control to Encounter Interstellar Objects” and published in Aerospace Research Central.
Artist’s illustration. ISOs like Oumuamua only come through once, making them difficult targets for rendezvous. Image Credit: NASA
“A human brain has many capabilities: talking, writing, etcetera,” Tsukamoto said in a press release. “Deep learning creates a brain specialized for one of these capabilities with a domain-specific knowledge. In this case, Neural-Rendezvous learns all the information it needs to encounter an ISO, while also considering the safety-critical, high-cost nature of space exploration.”
“Our key contribution is not just in designing the specialized brain, but in proving mathematically that it works,” Tsukamoto added. “For example, with a human brain we learn from experience how to navigate safely while driving. But what are the mathematics behind it? How do we know and how can we make sure we won’t hit anyone?”
The system is based on the “contraction theory for data-driven nonlinear control systems.” Contraction theory is a rigorous mathematical framework which can place limits on the effects of disturbances and uncertainties in complex linear systems. Basically, it can provide stability in a complex situation that changes nonlinearly over time.
The Neural-Rendezvous system uses available data to predict a spacecraft’s best actions to intercept an ISO. This complexity is necessary because ISOs are unbound and high-speed targets with poorly restrained trajectories.
“We’re trying to encounter an astronomical object that streaks through our solar system just once and we don’t want to miss the opportunity,” Tsukamoto said. “Even though we can approximate the dynamics of ISOs ahead of time, they still come with large state uncertainty because we cannot predict the timing of their visit. That’s a challenge.”
The Hubble space telescope captured this image of Comet 2l/Borisov at perihelion in December 2019. Image Credit: NASA, ESA, and D. Jewitt (UCLA)
ISOs only pass through the Solar System once. The usual method of observing an object like an asteroid or comet and determining its orbit doesn’t work. According to the researchers, it’s critical that ISO interceptors can “think” on their own.
“Unlike traditional approaches in which you design almost everything before you launch a spacecraft, to encounter an ISO, a spacecraft has to have something like a human brain, specifically designed for this mission, to fully respond to data onboard in real-time,” Tsukamoto said.
There’s no way to orbit an ISO. Oumuamua and Borisov were travelling at ~88 and 45 km/s relative to the Sun, so an intercepting spacecraft would need to travel at similar speeds. With our current technological level, a spacecraft would have to carry a prohibitively large volume of propellant to enter into orbit around one of these objects. Fast flybys are likely the only realistic mission architecture.
However, relying on a single spacecraft is like putting all your eggs in one basket. What if the spacecraft is unable to get a clear view of the ISO? Without a good look at the object, scientists won’t be able to learn much about its surface and composition. This has led some researchers to consider multiple spacecraft.
Tsukamoto worked with two other researchers on “a novel multi-spacecraft framework for locally maximizing information to be gained through ISO encounters.” Their work is presented in a separate paper titled “Information-Optimal Multi-Spacecraft Positioning for Interstellar Object Exploration.” Along with Tsukamoto, the other authors are Arna Bhardwaj and Shishir Bhatta. The authors presented it at the 2024 IEEE Aerospace Conference.
“Because of the speed and uncertainty, it’s challenging to obtain a clear view of an ISO during a flyby with 100 percent accuracy, even with Neural-Rendezvous,” Tsukamoto said. “Arna and Shishir wanted to show that Neural-Rendezvous could benefit from a multi-spacecraft concept.”
“Interstellar objects (ISOs), astronomical objects not gravitationally bound to the Sun, could present valuable opportunities to advance our understanding of the universe’s formation and composition,” the authors write in their paper. “In response to the unpredictable nature of their discoveries that inherently come with large and rapidly changing uncertainty in their state, this paper proposes a novel multi-spacecraft framework for locally maximizing information to be gained through ISO encounters with formal probabilistic guarantees.”
Their framework involves a swarm of spacecraft, called deputy spacecraft, and one designated as chief. The swarm would be located around an ellipsoid representing the space through which an ISO will travel. The ellipsoid consists of multiple points of interest (POIs) that would be covered collectively by the deputies and the chief employing the Neural-Rendezvous system. This method can maximize the information gained from the encounter. In simple terms, it guarantees multiple views of the ISO.
“Now we have an additional layer of decision-making during the ISO encounter,” Tsukamoto said. “How do you optimally position multiple spacecraft to maximize the information you can get out of it? Their solution was to distribute the spacecraft to visually cover the highly probable region of the ISO’s position, which is driven by Neural-Rendezvous.”
This simple drawing illustrates the deputy spacecraft in different positions in an ellipsoid an ISO is expected to pass through. Image Credit: Bhardwaj et al. 2025.
The number of spacecraft in the swarm would depend on the size of the uncertainty ellipsoid. The team ran simulations to examine the optimal number of spacecraft while keeping the cost down. With infinite resources, the swarm could be large enough to guarantee success. However, that’s not how things work.
In three trials, they determined that five spacecraft delivered the best results when balancing all factors.
This diagram shows the terminal positions of a five-spacecraft system and the POIs in view and not in view. Image Credit: Bhardwaj et al. 2025.
At the moment, Neural-Rendezvous is largely theoretical. However, the work done by Bhardwaj and Bhatta illustrates how it could be employed practically to maximum effect.
As unwitting messengers from other solar systems, ISOs have a scientific value that could surprise us. They could hold clues to how solar systems form and evolve that are found nowhere else. An autonomous swarm of spacecraft could help scientists collect these clues.