Description

This course provides a practical introduction to statistical orbit determination and space object tracking for space domain awareness, national security space, military space operations, and related civil and commercial applications. The course focuses on orbit-solving and tracking of noncooperative space objects using indirect measurements from radar, optical, infrared, passive RF, TDOA/FDOA, Doppler, and other distributed sensor systems.

Unlike cooperative spacecraft navigation, space object tracking often involves objects that do not provide trusted navigation solutions/data. Their orbits must instead be inferred from measurements that may be sparse, ambiguous, or biased. These conditions introduce challenges such as missed detections, false alarms, sensor biases, weak information observability, uncertain data association, and maneuvering targets.

The course covers the estimation and tracking concepts needed to address these problems while emphasizing physical intuition, uncertainty interpretation, operational constraints, and MATLAB-based examples. The goal is to help students understand not only how orbit determination and tracking algorithms work but also how to interpret their outputs and recognize their limitations in real tracking and decision-making scenarios.

What You Will Learn:

• Understand the space object tracking process from sensor observations to orbit determination, maintained tracks, and operational decisions.
• Build practical orbit determination problems using appropriate dynamics, measurements, coordinate frames, timing assumptions, and uncertainty models.
• Explain what different sensors can and cannot observe, including radar, optical, infrared, passive RF, Doppler, TDOA/FDOA, and distributed sensor systems.
• Estimate an orbit from limited data using initial orbit determination and batch least squares.
• Maintain a track over time using Kalman filtering, Extended Kalman filtering, Unscented Kalman filtering, and related sequential estimation concepts.
• Interpret tracking results using residuals, innovations, covariance, filter consistency checks, and track quality metrics.
• Recognize when a tracking solution may be unreliable because of weak geometry, model mismatch, bad initialization, sensor bias, missed detections, false alarms, or object maneuvers.
• Understand how track initiation/confirmation, sensor fusion, data association, maneuver detection, and sensor tasking support real space domain awareness operations.

Who Should Attend:

This course is intended for scientists, engineers, analysts, operators, and technical managers working in space, defense, government, FFRDC/UARC, contractor, commercial, and academic environments. It is especially relevant for personnel supporting space domain awareness, military space operations, Space Force missions, sensor systems, tracking, space control, or operational decision-making.

Course Outline:

1. The Space Object Tracking Problem (Understanding what makes space object tracking different from spacecraft navigation)

     This section introduces the full tracking problem: detecting objects, turning sensor data into measurements, estimating orbits, maintaining tracks, and supporting operational decisions. It also explains why noncooperative objects are harder to track than cooperative spacecraft and where uncertainty enters the problem.

 2. Orbit Dynamics and Force Modeling (Understanding how objects move in space and how modeling errors affect the estimated orbit)

   This section reviews the orbital dynamics needed for practical orbit determination, including two-body motion, numerical propagation, Earth gravity, atmospheric drag, solar radiation pressure, third-body effects, and the trade between simple and higher fidelity models. Emphasis is placed on knowing when model errors matter and how they show up in tracking results.

 3. Measurement Models and Sensor Geometry (Understanding what different sensors actually measure and what those measurements can tell you)

   This section covers common space tracking measurements from radar, optical, infrared, passive RF, Doppler, TDOA/FDOA, and distributed sensors. It also explains how sensor location, viewing geometry, observation timing, sensor biases, and measurement type affect what parts of the orbit can be estimated well.

 4. Initial Orbit Determination (Turning a small number of observations into an initial usable orbit estimate)

   This section introduces approaches for solving for an initial orbit estimate from limited tracking data. It discusses classical initial orbit determination methods, short-arc challenges, multiple possible orbit solutions, and techniques such as admissible region concepts for narrowing the set of physically reasonable possibilities.

 5. Batch Orbit Determination (Refining an orbit estimate using many observations at once)

   This section explains how a preliminary orbit is improved using a batch estimation process. Topics include weighted least squares, differential correction, residuals, measurement weighting, outlier handling, and estimating the uncertainty of the final orbit solution.

 6. Sequential Estimation and Kalman Filtering (Updating an orbit estimate as new measurements arrive)

   This section introduces the estimation concepts and terminology behind sequential tracking, including probability, uncertainty, covariance, prediction, measurement updates, process noise, and Kalman filtering. The focus is on building intuition for how filters combine a dynamic model with new sensor data over time.

 7. Nonlinear Orbit Estimation (Knowing how to handle nonlinearities that occur real orbit tracking)

   This section explains why orbit estimation is a mathematically nonlinear problem and how common nonlinear filters address that challenge. Topics include the Extended Kalman Filter, Unscented Kalman Filter, particle filter concepts, Gaussian mixture concepts, and when simple Gaussian assumptions may break down.

 8. Residuals, Covariance, and Filter Health (Knowing when an estimator is working and when it is lying to you)

   This section focuses on practical filter diagnostics: residuals, innovations, uncertainty bounds, consistency checks, covariance realism, and common warning signs of estimator trouble. Students learn how to recognize weak observability, poor initialization, bad noise assumptions, unmodeled biases, maneuvering objects, and filter divergence.

 9. Track Initiation, Maintenance, and Sensor Fusion (Managing a track over time, across gaps, and across multiple sensors)

   This section covers the lifecycle of a space object track: tentative tracks, confirmed tracks, track coast, reacquisition, deletion, and track quality assessment. It also introduces the practical issues involved in combining data from multiple sensors, handling sensor handover, and accounting for sensor registration errors.

 10. False Alarms, Gating, and Data Association (Deciding which measurements belong to which object and if they are real or not)

    This section introduces the tracking challenges caused by clutter, missed detections, ambiguous observations, and multiple nearby objects. Topics include validation gates, measurement-to-track assignment, nearest neighbor methods, global assignment, probabilistic association, multiple hypothesis concepts, and catalog correlation.

 11. Maneuvering Objects and Maneuver Detection (Recognizing when an object may have changed its orbit and how to handle those situations)

    This section discusses how maneuvers appear in tracking data and how they can be distinguished from sensor error, model error, or normal estimation uncertainty. Topics include impulsive and finite maneuvers, residual signatures, covariance growth, acceleration modeling, multiple model tracking, maneuver detection, and post-maneuver orbit recovery.

 12. Sensor Tasking and Advanced Orbital Regimes (Learning about sensor tasking approaches and exploring tracking in the cislunar regime)

    This section connects orbit estimation to sensor tasking. It introduces sensor visibility, revisit constraints, field of view limits, custody maintenance, information gain, and covariance-based tasking. It also discusses challenging orbital regimes such as GEO, highly elliptical, and cislunar.

Instructor(s):

Dr. Julian Brew is an aerospace engineer in the Space Exploration Sector at the Johns Hopkins University Applied Physics Laboratory, where he supports national security space (NSS), space domain awareness (SDA), and space control applications. His work focuses on astrodynamics, statistical mission planning, nonlinear space object tracking, autonomous spacecraft guidance, navigation, and control (GNC), and modeling and simulation. He has more than 13 years of experience supporting space-related research and engineering efforts for AFRL, NASA, and JHU/APL. Dr. Brew received his B.S., M.S., and Ph.D. in Aerospace Engineering from the Georgia Institute of Technology.