Spacecraft are trapped behind human control. When Ken Oguri supported small satellite and deep-space missions at Japan Aerospace Exploration Agency (JAXA), he saw how specific the navigation commands sometimes needed to be.

"I was creating very detailed commands for how many seconds to rotate the satellite body about which axis for how many degrees, then change the direction of rotation for so many seconds, then fire your engine for so many seconds," says Oguri, assistant professor of aeronautics and astronautics.

He though that this was not sustainable if humans intend to explore further into space — a feeling reinforced when he saw similar systems in place while conducting postdoctoral research at NASA’s Jet Propulsion Laboratory (JPL).

Since joining Purdue AAE faculty in January 2022, Oguri’s research group has been contributing to the Cislunar Initiative by developing solutions that will enable commercial work and science exploration in the area between the Earth and moon and beyond.

Making satellites more self-reliant is one of the main motivating factors in his work.

Ken Oguri, Assistant professor of Aeronautics and Astronautics

"It’s not only cumbersome, but almost impossible to do continuous interaction with satellites because of the limited number of antennas on the ground. Autonomous operation is necessary in the coming decade if we want to expand our space activity beyond Earth’s orbit and into cislunar space," Oguri says.

Oguri’s frameworks and algorithms could enable distant spacecraft to independently manage many of their own maneuvers, including orbit and attitude control, rendezvous and docking and sensing capabilities related to these maneuvers.

Safe systems in a chaotic environment

Oguri names three main factors for designing autonomous control algorithms: characterizing and modeling the nonlinear vehicle dynamics; estimating the orbital state of the vehicles as well as the properties of other objects; and planning maneuvers that anticipate vehicle faults.

"We want to make sure that our spacecraft is capable of doing perception and estimation of its relative position and velocity. It must also know how the other object is moving," Oguri says. "If that object is debris, we have to estimate the mass and also the moments of inertia to predict its motion. If we want to dock for servicing, we have to consider translational and rotational motion for both vehicles," Oguri says.

His team is even building in factors that consider the frequently changing lighting conditions in space — a critical piece for a perception system to understand its surroundings. These are the basic elements for planning an autonomous maneuver — but before they’ll be accepted, these systems must also expect the unexpected.

"We have to make sure our trajectory is not going to cause hazardous events, even under the circumstances of thrust shutdown," Oguri says. "This aspect is also going to be very important to building trust in autonomous systems in space."

A balancing act for future integration

In a separate but related project, Oguri is working on autonomous guidance algorithms that consider the varying gravitational forces between the Earth and moon. This is necessary for vehicles that will operate in the area of balanced gravity between them.

"There are some chaotic regions where these two forces are competing with each other. If we add small perturbations in one direction, it might go toward the moon or toward the Earth," Oguri says. "This requires more careful implementation of control and navigation."

Oguri hopes to leverage these nonlinear effects in his autonomous vehicle control research.

"Both of them are important for space autonomy. We need to carefully model nonlinear, sensitive, chaotic dynamics in space, and we also want to consider safe, autonomous control of six degrees of freedom vehicle motions," Oguri says. "For now, I’m starting in two different streams of projects, and I want to eventually merge them into one."