Background

Asteroids, particularly near earth object (NEO) asteroids are a major focus of US and international space based research. Asteroid missions were recommended in the 2003 Planetary Science Decadel Survey. At the Lunar and Planetary Science XXXIV (2003) conference further detail was provided with scientist proposing seismic studies and multi-instrument measurement pods as a way to get extensive geological, chemical and thermal measurements from multiple locations on the asteroid's surface. The most recent 2011 Planetary Science Decadel Survey recommended focus on analyzing trajectories to near earth asteroids.

 

tetrapod instrument

tetrapod instrument

 

The general engineering problem is to fly to an asteroid millions of kilometers away, place sensor multiple multi-instrument sensors on the surface and then gather data from those sensors. For many of the geological and seismic studies, knowledge of the sensor locations is important. In a terrestrial environment this could trivially be solved using GPS receivers on each instrument sensor. The luxury of an orbiting GPS constellation is not available in extra-terrestrial environments, however the principles could be used to create an ad-hoc pseudolite array, as was researched a few years ago in this lab. The hardware to achieve this comes with a significant weight and power penalty that would limit the number of instrument sensors that could be deployed.

RFID sensor

 At the other end of the spectrum of weight and power are entirely passive devices that rely on RFID technology for data communication. These devices weigh only grams and range from requiring milliwatts of power down to entirely passive devices. The main focus of the research in the ARL is to examine the feasibility of using RFID based sensors on an asteroid for measurement of sensor location, asteroid kinematics and asteroid dynamics.

Eros

Current Work

The starting point for the current research is the assumption that low cost SAW RFID-like sensors have been scattered over the body of the surface. An orbiting spacecraft with known inertial location interrogates these sensors both to gather scientific data, but also to build up a map of the sensor locations. SAW RFID devices have no accurate timing capabilities, so it assumed that time of flight information will be coarse at best. Frequency shift (via the doppler effect) is utilized to estimate a kinematic model of the motion of the asteroid. The asteroid kinematic state and sensor locations are jointly estimated in a SLAM-like problem using a range of Bayesian filters including Particle Filters, EKFs and UKFs.

The filters are being designed to explicitly handle the non-cartesian space that the kinematic attitude state of a rigid body exists in. Kinematic attitude is represented by quaternions and directional statistics are used to compute statistical moments.

Current error analyzes indicate that for non-pathological geometries, sensor locations can be estimated to a few meters precision on typically bodies. 

Sample Results

UKF filter shows good convergence for a tumbling asteroid

In the dual problem there is an axis ambiguity

Associated Personnel

Last modified Tue, 10 Feb, 2015 at 10:30