Seung H. Chung

Model-based programming of fault-aware systems

A wide range of sensor-rich, networked embedded systems are being created that must operate robustly for years in the face of novel failures by managing complex autonomic processes. These systems are being composed, for example, into vast networks of space, air, ground, and underwater vehicles. Our objective is to revolutionize the way in which we control these new artifacts by creating reactive model-based programming languages that enable everyday systems to reason intelligently and enable machines to explore other worlds. A model-based program is state and fault aware; it elevates the programming task to specifying intended state evolutions of a system. The programs executive automatically coordinates system interactions to achieve these states, entertaining known and potential failures, using models of its constituents and environment. At the executives core is a method, called CONFLICT-DIRECTED A*, which quickly prunes promising but infeasible solutions, using a form of one-shot learning. This approach has been demonstrated on a range of systems, including the National Aeronautics and Space Administrations Deep Space One probe. Model-based programming is being generalized to hybrid discrete-continuous systems and the coordination of networks of robotic vehicles.

Model learning for switching linear systems with autonomous mode transitions

We present a novel method for model learning in hybrid discrete-continuous systems. The approach uses approximate expectation-maximization to learn the maximum- likelihood parameters of a switching linear system. The approach extends previous work by 1) considering autonomous mode transitions, where the discrete transitions are conditioned on the continuous state, and 2) learning the effects of control inputs on the system. We evaluate the approach in simulation.

Ontology and modeling patterns for state-based behavior representation

This paper provides an approach to capture state-based behavior of elements, that is, the specification of their state evolution in time, and the interactions amongst them. Elements can be components (e.g., sensors, actuators) or environments, and are characterized by state variables that vary with time. The behaviors of these elements, as well as interactions among them are represented through constraints on state variables. This paper discusses the concepts and relationships introduced in this behavior ontology, and the modeling patterns associated with it. Two example cases are provided to illustrate their usage, as well as to demonstrate the flexibility and scalability of the behavior ontology: a simple flashlight electrical model and a more complex spacecraft model involving instruments, power and data behaviors. Finally, an implementation in a SysML profile is provided.

Practical application of model-based programming and state-based architecture to space missions

Innovative systems and software engineering solutions are required to meet the increasingly challenging demands of deep-space robotic missions. While recent advances in the development of integrated systems and software engineering approaches have begun to address some of these issues, these methods are still at the core highly manual and, therefore, error-prone. This paper describes a task aimed at infusing MITs model-based executive, Titan, into JPLs Mission Data System (MDS), a unified state-based architecture, systems engineering process, and supporting software framework. Results of the task are presented, including a discussion of the benefits and challenges associated with integrating mature model-based programming techniques and technologies into a rigorously-defined domain specific architecture

Signicance of Decomposition in Spacecraft Autonomy

Unlike traditional ight software used in spacecraft, many autonomy software use search algorithms that solve problems with NP complexity or worse. Without exhaustively testing the state space of search algorithms, we can only conclude that the worst case memory or time performance will be exponential or worse. This, however, is unacceptable for missions that are time critical and memory limited. The arti cial intelligence community has also been concerned with the intractability of AI algorithms. As result, a set of decomposition techniques, namely constraint decomposition and causal order decomposition, have been developed to address the issue. Decomposition techniques use a divide-and-conquer approach to divide a problem based on the properties of the problem into a set of simpler subproblems. By decomposing the problem, we are able to determine a tighter bound on the time and memory required to solve the problem and use the decomposition to solve the problem within the guaranteed time and memory. Furthermore, this approach can help design and verify autonomy software. In this paper we review two important decomposition techniques, constraint decomposition and causal order decomposition and their uses within spacecraft autonomy.