SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Space ROS is a matured version of ROS 2 using flight software engineering practices to facilitate mission- and safety-critical applications. Our goal is to reduce the barrier to maturation and infusion of software developed using the popular ROS framework.
Runtime Assurance (RTA) is a design-time architecture for safety-critical systems where an internal monitor acts upon detecting a violation of a property. The simplex architecture is an instance of RTA, where the action taken is to hand control of the overall system to a trusted controller when an untrusted one violates a safety property. Simplex RTA is emerging as a method for allowing AI/ML and other unverified software to be integrated into safety-critical applications like aircraft. To this end, the American Society for Testing and Materials (ASTM) and NASA have each published guidelines on the use of RTA in such systems. In the simplex RTA framework, a system has an advanced controller (AC) and a reversionary controller (RC). The system is allowed to operate with the AC until a runtime monitor detects that some property has been violated and then the RC takes over. Assuming that the sample rate of the monitor will detect improper functioning with enough time for the RC to correct the impending problem, and that the RC is trusted, the system will operate as intended. This use of the simplex RTA framework can allow for the integration of untrusted, but possibly more performant, controllers in a safe way. This paper presents a formalization of a simplex RTA framework in the Prototype Verification System (PVS) theorem prover using an embedding of differential dynamic logic (DDL) called Plaidypvs. A novel feature of this framework is that it can be instantiated at different levels of abstraction. This feature allows for the formal verification of a system with an untrusted black box component, such as an AI/ML controller. This paper does not address the many difficulties in deploying RTA in an industrial-level system. Instead, the focus is on the formal verification of the simplex RTA framework in the language of hybrid programs. Hybrid programs are programs that include both discrete and continuous dynamics and can be used to model complex cyber-physical systems. Plaidypvs is a tool that enables formalization of hybrid programs in the PVS theorem prover. Plaidypvs enables the verification of the general simplex RTA framework and then, by specializing some components of the hybrid program, verifying instances of the framework while treating the untrusted component as a black box. A selection of Unmanned Aircraft Systems (UAS) operations are shown as instances of the general RTA framework in PVS. This offers the benefit of design time verification of relevant safety properties to the system, and it also gives requirements on the sample rate of sensors that determine the time interval in which the ‘switch’ property of the RTA framework is checked.
New concepts for automating air traffic management functions at small non-towered airports raise serious safety issues associated with the software implementations and their underlying key algorithms. The criticality of such software systems necessitates that strong guarantees of the safety be developed for them. In this paper we present a formal method for modeling and verifying such systems using the PVS theorem proving system. The method is demonstrated on a preliminary concept of operation for the Small Aircraft Transportation System (SATS) project at NASA Langley.
NASA's Johnson Space Center (JSC) / Flight Design and Dynamics Division (DM) has prototyped the use of Open Source middleware technology for building its next generation spacecraft mission support system. This is part of a larger initiative to use open standards and open source software as building blocks for future mission and safety critical systems. JSC is hoping to leverage standardized enterprise architectures, such as Java EE, so that its internal software development efforts can be focused on the core aspects of their problem domain. This presentation will outline the design and implementation of the Trajectory system and the lessons learned during the exercise.
System safety, an integral component in software development, often poses a challenge to engineers designing computer-based systems. While the relaxed constraints on software design allow for increased power and flexibility, this flexibility introduces more possibilities for error. As a result, system engineers must identify the design constraints necessary to maintain safety and ensure that the system and software design enforces them. Safeware Engineering Corporation, of Seattle, Washington, provides the information, tools, and techniques to accomplish this task with its Specification Tools and Requirements Methodology (SpecTRM). NASA assisted in developing this engineering toolset by awarding the company several Small Business Innovation Research (SBIR) contracts with Ames Research Center and Langley Research Center. The technology benefits NASA through its applications for Space Station rendezvous and docking. SpecTRM aids system and software engineers in developing specifications for large, complex safety critical systems. The product enables engineers to find errors early in development so that they can be fixed with the lowest cost and impact on the system design. SpecTRM traces both the requirements and design rationale (including safety constraints) throughout the system design and documentation, allowing engineers to build required system properties into the design from the beginning, rather than emphasizing assessment at the end of the development process when changes are limited and costly.System safety, an integral component in software development, often poses a challenge to engineers designing computer-based systems. While the relaxed constraints on software design allow for increased power and flexibility, this flexibility introduces more possibilities for error. As a result, system engineers must identify the design constraints necessary to maintain safety and ensure that the system and software design enforces them. Safeware Engineering Corporation, of Seattle, Washington, provides the information, tools, and techniques to accomplish this task with its Specification Tools and Requirements Methodology (SpecTRM). NASA assisted in developing this engineering toolset by awarding the company several Small Business Innovation Research (SBIR) contracts with Ames Research Center and Langley Research Center. The technology benefits NASA through its applications for Space Station rendezvous and docking. SpecTRM aids system and software engineers in developing specifications for large, complex safety critical systems. The product enables engineers to find errors early in development so that they can be fixed with the lowest cost and impact on the system design. SpecTRM traces both the requirements and design rationale (including safety constraints) throughout the system design and documentation, allowing engineers to build required system properties into the design from the beginning, rather than emphasizing assessment at the end of the development process when changes are limited and costly.
Documented here is the recent effort of the MISSION Kernel Team on an Extra-Vehicular Activity Control System (EVACS) simulation with nested transactions. The team has implemented the EVACS simulation along with a design for nested transactions. The EVACS simulation is a project wide aid to exploring Mission and Safety Critical (MASC) applications and their support software. For this effort it served as a trial scenario for demonstrating nested transactions and exercising the transaction support design. The EVACS simulation is a simulation of some aspects of the Extra-Vehicular Activity Control System, in particular, just the selection of communication frequencies. Its current definition is quite narrow, serving only as a starting point for prototyping purposes. (EVACS itself may be supplanted in a larger scenario of a lunar outpost with astronauts and a lunar rover.) Initially the simulation of frequency selection was written without consideration of nested transactions. This scenario was then modified to embed its processing in nested transactions. To simplify the prototyping effort, only two aspects of the general design for transaction support have been implemented: the basic architecture and state recovery. The simulation has been implemented in the programming language Smalltalk. It consists of three components: (1) a simulation support code which provides the framework for initiating, interacting and tracing the system; (2) the EVACS application code itself, including its calls upon nested transaction support; and (3) a transaction support code which implements the logic necessary for nested transactions. Each of these components deserves further description, but for now only the transaction support is discussed.
An external review of the Integrated Technology Plan for the Civil Space Program is presented. This review is specifically concerned with the Software Engineering Program. The goals of the Software Engineering Program are as follows: (1) improve NASA's ability to manage development, operation, and maintenance of complex software systems; (2) decrease NASA's cost and risk in engineering complex software systems; and (3) provide technology to assure safety and reliability of software in mission critical applications.
Adaptive control technologies that incorporate learning algorithms have been proposed to enable autonomous flight control and to maintain vehicle performance in the face of unknown, changing, or poorly defined operating environments [1-2]. At the present time, however, it is unknown how adaptive algorithms can be routinely verified, validated, and certified for use in safety-critical applications. Rigorous methods for adaptive software verification end validation must be developed to ensure that. the control software functions as required and is highly safe and reliable. A large gap appears to exist between the point at which control system designers feel the verification process is complete, and when FAA certification officials agree it is complete. Certification of adaptive flight control software verification is complicated by the use of learning algorithms (e.g., neural networks) and degrees of system non-determinism. Of course, analytical efforts must be made in the verification process to place guarantees on learning algorithm stability, rate of convergence, and convergence accuracy. However, to satisfy FAA certification requirements, it must be demonstrated that the adaptive flight control system is also able to fail and still allow the aircraft to be flown safely or to land, while at the same time providing a means of crew notification of the (impending) failure. It was for this purpose that the NASA Ames Confidence Tool was developed [3]. This paper presents the Confidence Tool as a means of providing in-flight software assurance monitoring of an adaptive flight control system. The paper will present the data obtained from flight testing the tool on a specially modified F-15 aircraft designed to simulate loss of flight control faces.
This paper describes a project appling formal methods to a portion of the shuttle on-orbit digital autopilot (DAP). Three objectives of the project were to: demonstrate the use of formal methods on a shuttle application, facilitate the incorporation and validation of new requirements for the system, and verify the safety-critical properties to be exhibited by the software.
For mission and safety critical systems it is necessary to: improve definition, evolution and sustenance techniques; lower development and maintenance costs; support safe, timely and affordable system modifications; and support fault tolerance and survivability. The goal of the MISSION project is to lay the foundation for a new generation of integrated systems software providing a unified infrastructure for mission and safety critical applications and systems. This will involve the definition of a common, modular target architecture and a supporting infrastructure.
As computers become more widely used, and in particular as they become used in more safety critical applications, the reliability of the computer system and its software becomes more important. There is also an increasing need for high levels of reliability in applications involving very large numbers of inexpensive units where recall of the units would be disproportionately expensive. The nature of faults and the assumptions made by different approaches to correct operation are considered. The recovery block approach is described and a probabilistic analysis of its effectiveness, with and without correlated design errors is provided. Mechanisms for generating acceptance tests from specifications, and for providing recovery in the presence of asynchrony, are described. An analysis of, and design for, the provision of recovery blocks in the microprogram of the Bendix BDX930 processor is provided. An example of the use of recovery blocks in a simple operating system is also provided.
This paper describes experience using Obstacle Analysis to identify contingency requirements on an unpiloted aerial vehicle. A contingency is an operational anomaly, and may or may not involve component failure. The challenges to this effort were: ( I ) rapid evolution of the system while operational, (2) incremental autonomy as capabilities were transferred from ground control to software control and (3) the eventual safety-criticality of such systems as they begin to fly over populated areas. The results reported here are preliminary but show that Obstacle Analysis helped (1) identify new contingencies that appeared as autonomy increased; (2) identify new alternatives for handling both previously known and new contingencies; and (3) investigate the continued validity of existing software requirements for contingency handling. Since many mobile, intelligent systems are built using a development process that poses the same challenges, the results appear to have applicability to other similar systems.
Operational requirements of safety-critical systems are often written in restricted specification logics. These restricted logics are amenable to automated analysis techniques such as model-checking, but are not rich enough to express complex requirements of unmanned systems. This short paper advocates for the use of expressive logics, such as higher-order logic, to specify the complex operational requirements and safety properties of unmanned systems. These rich logics are less amenable to automation and, hence, require the use of interactive theorem proving techniques. However, these logics support the formal verification of complex requirements such as those involving the physical environment. Moreover, these logics enable validation techniques that increase con dence in the correctness of numerically intensive software. These features result in highly-assured software that may be easier to certify. The feasibility of this approach is illustrated with examples drawn for NASA's unmanned aircraft systems.
Modeling guidelines constitute one of the fundamental cornerstones for Model Based Development. Their relevance is essential when dealing with code generation in the safety-critical domain. This article presents the experience of a railway signaling systems manufacturer on this issue. Introduction of Model-Based Development (MBD) and code generation in the industrial safety-critical sector created a crucial paradigm shift in the development process of dependable systems. While traditional software development focuses on the code, with MBD practices the focus shifts to model abstractions. The change has fundamental implications for safety-critical systems, which still need to guarantee a high degree of confidence also at code level. Usage of the Simulink/Stateflow platform for modeling, which is a de facto standard in control software development, does not ensure by itself production of high-quality dependable code. This issue has been addressed by companies through the definition of modeling rules imposing restrictions on the usage of design tools components, in order to enable production of qualified code. The MAAB Control Algorithm Modeling Guidelines (MathWorks Automotive Advisory Board)[3] is a well established set of publicly available rules for modeling with Simulink/Stateflow. This set of recommendations has been developed by a group of OEMs and suppliers of the automotive sector with the objective of enforcing and easing the usage of the MathWorks tools within the automotive industry. The guidelines have been published in 2001 and afterwords revisited in 2007 in order to integrate some additional rules developed by the Japanese division of MAAB [5]. The scope of the current edition of the guidelines ranges from model maintainability and readability to code generation issues. The rules are conceived as a reference baseline and therefore they need to be tailored to comply with the characteristics of each industrial context. Customization of these recommendations has been performed for the automotive control systems domain in order to enforce code generation [7]. The MAAB guidelines have been found profitable also in the aerospace/avionics sector [1] and they have been adopted by the MathWorks Aerospace Leadership Council (MALC). General Electric Transportation Systems (GETS) is a well known railway signaling systems manufacturer leading in Automatic Train Protection (ATP) systems technology. Inside an effort of adopting formal methods within its own development process, GETS decided to introduce system modeling by means of the MathWorks tools [2], and in 2008 chose to move to code generation. This article reports the experience performed by GETS in developing its own modeling standard through customizing the MAAB rules for the railway signaling domain and shows the result of this experience with a successful product development story.
NASA IV&V was established in 1993 to improve safety and cost-effectiveness of mission critical software. Since its inception the tools and strategies employed by IV&V have evolved. This paper examines how lessons learned from the Phoenix project were developed and applied to the GRAIL project. Shortly after selection, the GRAIL project initiated a review of the issues documented by IV&V for Phoenix. The motivation was twofold: the learn as much as possible about the types of issues that arose from the flight software product line slated for use on GRAIL, and to identify opportunities for improving the effectiveness of IV&V on GRAIL. The IV&V Facility provided a database dump containing 893 issues. These were categorized into 16 bins, and then analyzed according to whether the project responded by changing the affected artifacts or using as-is. The results of this analysis were compared to a similar assessment of post-launch anomalies documented by the project. Results of the analysis were discussed with the IV&V team assigned to GRAIL. These discussions led to changes in the way both the project and IV&V approached the IV&V task, and improved the efficiency of the activity.
Automated code generators (ACG) are tools that convert a (higher-level) model of a software (sub-)system into executable code without the necessity for a developer to actually implement the code. Although both commercially supported and in-house tools have been used in many industrial applications, little data exists on how these tools are used in safety-critical domains (e.g., spacecraft, aircraft, automotive, nuclear). The aims of the survey, therefore, were threefold: 1) to determine if code generation is primarily used as a tool for prototyping, including design exploration and simulation, or for fiight/production code; 2) to determine the verification issues with code generators relating, in particular, to qualification and certification in safety-critical domains; and 3) to determine perceived gaps in functionality of existing tools.
Within the past decade, software has become an increasingly common element in computing systems. In particular, the role of software used in the aerospace industry, especially in life- or safety-critical applications, is rapidly expanding. This intensifies the need to use effective techniques for achieving and verifying the reliability of avionics software. Although certain software development processes and techniques are mandated by government regulating agencies, no one methodology has been shown to consistently produce reliable software. The knowledge base for designing reliable software simply has not reached the maturity of its hardware counterpart. In an effort to increase our understanding of software, the Langley Research Center conducted a series of experiments over 15 years with the goal of understanding why and how software fails. As part of this program, the effectiveness of current industry standards for the development of avionics is being investigated. This study involves the generation of a controlled environment to conduct scientific experiments on software processes.
Software health management (SWHM) techniques complement the rigorous verification and validation processes that are applied to safety-critical systems prior to their deployment. These techniques are used to monitor deployed software in its execution environment, serving as the last line of defense against the effects of a critical fault. SWHM monitors use information from the specification and implementation of the monitored software to detect violations, predict possible failures, and help the system recover from faults. Changes to the monitored software, such as adding new functionality or fixing defects, therefore, have the potential to impact the correctness of both the monitored software and the SWHM monitor. In this work, we describe how the results of a software change impact analysis technique, Directed Incremental Symbolic Execution (DiSE), can be applied to monitored software to identify the potential impact of the changes on the SWHM monitor software. The results of DiSE can then be used by other analysis techniques, e.g., testing, debugging, to help preserve and improve the integrity of the SWHM monitor as the monitored software evolves.