“When it first starts flying and undocks, the reference frame
would be with respect to the mothership,” explains DuPuis. “As
the vehicle makes its way to the asteroid, to map, land, and take
samples, the vehicle needs to operate in a reference frame with
respect to the asteroid.”
Another reason why the process is so complex is because the
researchers can’t use many sensors that flight vehicles use on
earth, such as an altimeter, magnetometer, or GPS. “We have to
come up with other ways to estimate positions and velocities,”
says DuPuis. One of the ways the team does this is with
image-based navigation and ranging technologies, although the
algorithms required to process this data are difficult to process in
To control the vehicle’s various states, the team is
implementing non-linear controller techniques which are more
capable than classic proportional-integral-derivative (PID)
controllers, a common control architecture for liner systems.
However, while non-linear control is a more robust method, it is
also more computationally intensive.
“One of the things that the controller has to do is manage
all of the different thrusters around the vehicle to produce the
desired states,” explains DuPuis. But, the vehicle’s thrust can’t
be throttled in the same way as a car’s engine. Instead, to keep
reliability high, the thrusters feature solenoid values that simply
open and close to control the total impulse on the vehicle.
According to DuPuis, the amount of thrust produced is relatively
constant when the solenoids are open; however, the system
required to pulse the thrusters, both from a software, sensor and
a hardware standpoint, presents challenges.
One of the challenges on the hardware side is the life of the
solenoid itself and being able to control the vehicle in ways
that maximize its lifespan. “Because the design lifecycle of
the solenoid valve is life-limited,” says DuPuis, “you have to
understand the limitations and manage them accordingly.”
Another challenge from a hardware perspective has been
developing the systems needed to secure the vehicles to the
surface. While the system may not be required for the vehicle
operating in partial gravity, in zero to mili-gravity, if the craft isn’t
anchored to the surface, the reaction force from drilling would
push the spacecraft away.
“We saw that happen on the Rosetta mission with the Philae
lander when it bounced several times and landed under a cliff
face,” DuPuis notes. “In dramatic fashion, it highlighted the risk
of trying to land on a low-gravity body … [It was] an excellent
example of why you might want to use one or more low-
cost marsupial spacecraft and not risk the larger mothership
According to DuPuis, these risks can be mitigated in two
different ways. One is through the development of the grappling,
or anchoring system, the other is with a very low reaction force
The system features pneumatic samplers that may simply use a
small puff of gas to kick up a sample for capture. A low reaction
rotary-percussive drill may also be used.
“In all cases we have industry and academic partners that work
with us on the development of these technologies,” says DuPuis.
To date, the team has only been able to develop and test the
Asteroid Free Flyer’s attitude control. The project’s next phase
is developing the technology in the lab to demonstrate the
vehicle’s ability to translate.
“The greatest engineering challenge is to demonstrate
technology that’s intended to be used in space here on earth,
which is a very different environment,” explains DuPuis.
To overcome these challenges, systems have been developed
in the past by organizations within NASA and other companies,
such as an active gravity offloading system.
“The vehicle would be suspended, moving around with an
active suspension system that is continuously offloading the
vehicles weight so that it can move with small thrust impulses,”
While the Asteroid Free Flyer prototype uses cold gas
thrusters, the Mars Free Flyer prototype still uses air-breathing
thrusters as analogs or surrogates for the actual thrusters. The
team hopes to implement the ISRU cold gas thruster system on
the Mars Free Flyer prototype in the coming months.
For both systems, the team hopes to further refine the
image-based navigation system and its capabilities, as its
advancement will not only benefit the Extreme Access Flyer
project, but also autonomous search and rescue operations in
hazardous or hard-to-reach areas.
One of the particularly applicable implementations of vision-aided navigation is simultaneous localization and mapping
(SLAM), a process in which robots need to understand
obstacles in real time in order to plan a path, “Much in the
same way you and I would, if we walked into a room for the first
time,” explains DuPuis. This technology could also be applied
here on earth to robotic vehicle operations in autonomous
search and rescue operations and other useful tasks.
However, the project’s original goal and the gravity of
developing sustainable space flight is not lost on the team.
“We’re trying to find resources in extreme environments
using the robotic prospector,” says Robert P. Mueller, senior
technologist, NASA Kennedy Space Center Swamp Works.
The Asteroid Free Flyer prototype
in a testing gimbal.