Recent Trends and Advances in Model Based Systems Engineering

Digital and model-based engineering envisions a future where software systems are intricately involved in systems engineering and engineering design efforts. Recent advances in the field of software engineering have the potential to enable more flexible, reconfigurable, and updateable systems for engineering applications. This paper introduces an information system reference architecture for digital and model-based engineering activities based on modern web-based architectural styles. An application case explains how the reference architecture shaped the implementation of the Tradespace Analysis Tool for Constellations (TAT-C) Knowledge Base, a software component for space systems engineering that maintains a resource library conforming to common object schemas. Database, back-end, and front-end software components serve as architectural layers connected by simple information protocols based on semantic linked data models for improved interoperability.

The construction of megaprojects has consistently demonstrated challenges for project managers in regard to meeting cost, schedule, and performance requirements. Megaproject construction challenges are commonplace within the nuclear industry with many active projects in the United States failing to meet cost and schedule efforts by significant margins. Currently, nuclear engineering teams operate in siloed tools and disparate teams where connections across design, procurement, and construction systems are translated manually or over brittle point-to-point integrations. The manual nature of data exchange increases the risk of silent errors in the reactor design, with each silent error cascading across the design. These cascading errors lead to uncontrollable risk during construction, resulting in significant delays and cost overruns. Additionally, due to the desire to reduce schedule and avoid escalation, construction is often begun prior to full design maturity. Digital engineering (DE) embodies a deliberate transformational approach to the manner in which systems are designed, engineered, constructed, operated, maintained, and retired. DoD defines DE as “an integrated digital approach that uses authoritative sources of system data and models as a continuum across disciplines to support lifecycle activities from concept through disposal” (U.S. Department of Defense, Digital Engineering Strategy, Washington, DC, June 2018). This paper describes the ontologies (data model), tool architectures, data exchange, and process to transform engineering teams to a new digital engineering ecosystem.

Healthcare policy evaluation is a time-consuming, challenging process due to the complexity of the US healthcare system which is comprised of both public and private payers; a variety of healthcare suppliers including doctors, medical device companies, and pharmacies; and patients from different insurance coverages and socioeconomic backgrounds. Systems engineering processes are intended for complex systems and are ideal for addressing healthcare policy. Specifically, model-based systems engineering (MBSE) is used to increase traceability with its model-centric approach and can be used to increase understanding of the healthcare system. In this paper, we attempt to exploit digital twin philosophy of MBSE to understand a US healthcare system as a complex system. We focus our efforts in building a digital doppelgänger which reflects most aspects of the healthcare systems digitally, but is not an exact digital twin. The doppelgänger helps navigate around the medical privacy laws of the US healthcare system and runs some analysis on healthcare policy.

Model-based systems engineering (MBSE) requires greater investment than traditional systems engineering in the early phases of the system life cycle. Program management justifies this additional investment by arguing that such investments can be expected to produce continuous gains across later phases of system life cycle resulting from early detection of defects, risk reduction, improved communication, superior integration of the supply chain, product line definition, and enhanced traceability. Since systems evolve over the system life cycle, system models need to be updated continually to reflect the state of the system and realize value. However, in systems engineering organizations today, there is a tendency to reallocate modeling resources to other projects once initial modeling is completed on a particular project. This practice results in a resource gap which impedes the continuous update of system models through later phases of the system life cycle. This paper presents how digital twin technology can be exploited to address model updates throughout the MBSE life cycle. This paper also presents preliminary results from prototyping and experiments with a digital twin including data collection from a physical system operating in the real world to update the digital twin model. This paper shows how operational analysis and modeling can be enhanced by leveraging the digital twin construct.

Increasing system complexity has been a driving force for the development of systems engineering methodologies. One of these methodologies is a concurrent engineering process known as set-based design (SBD). In support of ongoing SBD research, our team conducted a structured literature review to ascertain the current state of SBD research. This paper provides the results and analysis from the review of relevant SBD methodologies attempting to identify methodologies combining two or more systems analysis methodologies with SBD. The purpose of this research is to identify potential research opportunities by identifying underrepresented SBD methodologies in the literature. We specifically focus on applications combining SBD and model-based systems engineering (MBSE) with requirements development, decision analysis, risk analysis, affordability analysis, resilience, and complexity analysis applications. Our findings identified several potential research opportunities in these application areas, as well as additional opportunities for combining SBD and MBSE with systems architecting, uncertainty analysis, and intelligent adversary analysis.

Digital modernization is a relatively new concept that the Department of Defense (DoD) has embraced and recommended for implementation across the services. We explain the concepts behind digital modernization using a systems approach to avoid confusion and promote clarity. Many other definitions of these concepts include unnecessary combination of descriptions, goals, and methods. We describe digital modernization in terms of its purpose, elements, and interconnections. Additionally, we offer new simpler definitions of the digital thread, digital system model, and digital twin concepts. To conclude we propose ways in which digital modernization could be implemented in systems engineering practice.

Model curation can be thought of as a hallmark of digital maturity, signifying that models are highly valuable assets of the enterprise used as a trusted basis for engineering decisions. As enterprises begin to develop large model repositories, model credibility becomes a central concern. Recent studies show the decision to use, reuse, and repurpose models is contingent on the model consumer’s perception of validity and trustworthiness of the model. This paper discusses an investigation of selected foundational works on credibility of models, simulations, and websites as part of a larger research effort on model curation. The objective is to leverage findings and strategies from prior work and identify useful heuristics that can inform model credibility within the context of model curation.

The evaluation of a system’s architecture is an essential process within the systems engineering lifecycle. Commercially available model-based systems engineering (MBSE) tools, when combined with standards-based architecture modeling languages, provide a means through which architecture information can be expressed graphically and formally in a machine-readable format; this format can be leveraged in order to improve the system architecture evaluation process. The authors propose an automated, repeatable method for detecting patterns of interest embedded within an MBSE model. This novel method uses a heuristically guided set of similarity measures that depend on textual and graphical content of a model. The proposed method has been implemented for architectures developed in IBM’s Rational Rhapsody, and No Magic Inc.’s MagicDraw, and has proven to be able to identify six well-established patterns: Adapter, Bridge, Composite, Façade, Observer, and Proxy. This automation has the potential to produce cost and time savings for the evaluation process and to add an additional degree of rigor and completeness to an architecture evaluation.

CubeSats have the demonstrated potential to contribute to commercial, scientific, and government applications in remote sensing, communications, navigation, and research. Despite significant research into improving CubeSat operational efficiency, there remains one fundamental limitation of CubeSats for EO imaging applications: the small lenses and short focal lengths result in imagery with low spatial resolution. This paper reviews the previous research on super-resolution techniques and proposes potential applications of super-resolution to CubeSat EO imagery.

A honeypot system can play a significant role in exposing cybercrimes and maintaining reliable cybersecurity. Markov decision process (MDP) is an important method in systems engineering research and machine learning. The data analytics of a honeypot system based on an MDP model is conducted using R language and its functions in this paper. Specifically, data analytics over a finite planning horizon (for an undiscounted MDP and a discounted MDP) and an infinite planning horizon (for a discounted MDP) is performed, respectively. Results obtained using four kinds of algorithms (value iteration, policy iteration, linear programming, and Q-learning) are compared to check the validity of the MDP model. The simulation of expected total rewards for various states is implemented using various transition probability parameters and various transition reward parameters.

Complex systems that continuously interact with dynamic uncertain environments need the ability to adapt their decision-making based on observed outcomes of their decisions and actions. Traditional deterministic modeling approaches are invariably inadequate for modeling systems whose models are not fully known initially. For such systems, we need the ability to start with an incomplete model and then progressively complete the model with observations made along the way. To address this problem type, we propose an extendable-partially observable Markov decision process (extendable-POMDP) to model the system’s state space and decision-making in the presence of uncertainties. The extendable-POMDP model is able to account for unknown-unknowns by incorporating “new hidden states” that result in expanding the model state space which in turn extends the associated probability distributions. This paper provides an online algorithm for solving a POMDP model by employing a heuristic search algorithm that estimates long-term rewards in a finite-horizon look-ahead in a sense-plan-act cycle. Heuristics are employed in model definition, expansion, and online look-ahead search to contain the otherwise inevitable computational complexity arising from state-space explosion.

Sociotechnical systems (STSs) such as infrastructure management systems operate under highly dynamic, contextual, and environmental conditions; therefore they depend on specifically trained group of individuals known as Controllers for their safety-critical decision-making activities. The dependency of STS on human decision-makers introduces additional complexity to the system due to the intertwined social and technical factors that influence the operational decision-making process. While the role allocated to autonomous decision-making units in STSs is rapidly increasing, hard-to-estimate and the inherently unique nature of safety-critical situations render high levels of automation infeasible and require manual control in many instances. In this paper, we investigate real-world operational data from INFRABEL (Belgian Railways) and utilize data analytics techniques to understand how Controllers behave during adverse operational conditions. The identification and evaluation of adverse operational instances can also support the design of future decision support or automation tools. To achieve this, we first provide a brief discussion of social and technical factors that influence the Controllers and their decision-making process. We then introduce robust principal component analysis that is a rigorous data analytics technique to identify influential observations (leverage points and outliers). We finally provide a brief discussion of how operations during adverse conditions differ from nominal conditions. We observe that adverse operational conditions in our application typically occur in 5% of the observations. The proposed approach will be implemented at INFRABEL, the Belgian Railways.

Understanding system vulnerabilities to risk factors during operation is essential for developing dependable systems. By implication, assessing in-use risk factors requires monitoring system parameters that contribute to making probabilistic inferences. We argue, however, that naïve use of statistical data without regard to causality can yield surprising and often erroneous risk predictions. Making reliable risk predictions is further complicated by lack of full awareness of system states and the existence of unobservable parameters in complex systems. Overly conservative risk assessment leads to increased life-cycle cost and reduced system availability resulting from overly aggressive preventive maintenance or replenishment strategies, while overly optimistic risk assessment can lead to even higher life-cycle cost and potential harm when otherwise preventable failures occur. This paper discusses a causality-aware, dynamic risk assessment model based on hidden Markov model construct. This model employs the concept of hidden system states that account for otherwise unexplainable observations. The model is continuously evaluated during system operation and updated when new observations warrant reevaluation.

With increasing connectivity and digitalization, systems continue to become increasingly more complex. To meet this challenge, the model-based systems engineering community has begun exploring the use of ontologies to scope the modeling effort and demonstrate the value of MBSE without resorting to full-blown modeling. To this end, this paper presents a minimum viable model (MVM) approach to system modeling. In the MVM approach, a system model with the requisite structure and just enough semantics is created to resolve semantic inconsistencies in the model, achieve interoperability, and answer a few key questions at the right level of detail posed by stakeholders from the systems acquisition and engineering communities. The larger intent is to have potential customers buy into the viability of an ontology-enabled approach to MBSE.

This paper explores the critical issue of temporal modeling and reasoning for successful systems of systems (SoS) architecting and operations. The increasing complexity of missions results into needs to leverage capabilities of constituent systems (CSs) in multiple domains, distributed geographically and temporally (different time zones) too. This introduces issues capable of hindering mission success including clock drifts and synchronization as well as communication delays. To assure correctness of SoS functionality in the face of these challenges, we develop and introduce a new ontological framework for modeling and time-based reasoning in SoS. Knowledge representation of time and temporal semantics in SoS modeling are discussed with a focus on the central role description logics (DL) and interval-based time semantics play in the development of the new framework. The latter consists of a DL-backed theoretical foundation providing formalisms to ontological models encapsulating temporal knowledge on top of which time-based modeling and reasoning applications for SoS can be built. A prototype implementation with a military-directed SoS has been illustrated and is currently under development.

Adaptive systems are class of systems that change their behavior in response to external or internal disrupting events. These kinds of system are of interest to government, industry, and research community. An instance of an adaptive system is a multi-UAV swarm operating in open, dynamic environment. Such systems are employed to carry out missions such as search and rescue and disaster relief. Developing adaptive systems requires a hardware-software testbed to explore system behavior and make adjustments to achieve desired behaviors. This paper presents an ontology-enabled approach for developing integrated hardware-software testbed for engineering adaptive systems. Multi-UAV operation is used as an illustrative example of an adaptive system that stands to benefit from a hardware-software experimentation testbed.

System reconfiguration (SR) is essential in system management, as it is an enabler for system flexibility and adaptability, attendant ilities being reliability, availability, maintainability, testability, safety, and reuse of system entities and technologies. Within current industrial practice, the development of reconfiguration tools is a real challenge. The development of these tools demand clear identification of reconfiguration data. In this paper, key concepts of the reconfiguration process, and relations among them, are represented in the form of the OSysRec ontology. These concepts are applied to the integrated modular avionics case study to test the proposed ontology within the aerospace domain.

Digital engineering (DE) provides an opportunity to reduce engineering design rework. However, this potential depends significantly on the approach used in exploring the design tradespace. A classical approach, using a traditional point-based design (PBD), has the likelihood of creating engineering rework even within the context of digital engineering. To address limitations of PBD, several researchers have proposed the use of set-based design (SBD). However, there is no formal definition of SBD and there is limited guidance on how to effectively implement it in a DE environment. To address such concerns, a literature review was conducted around the following questions: (1) What is the current state of SBD application – approaches and models? (2) What are the strengths and limitations to these approaches and models? (3) How is knowledge developed, captured, and reused to cause convergence of sets? (4) What DE tools were recommended/used to enable SBD?

Large-scale systems engineering projects may be construed essentially as multi-agent problems wherein decisions are made by several people (stakeholders, managers, designers, etc.) across the organizational hierarchy. All agents in an enterprise possess knowledge in one form or the other, be it the knowledge gained from requirements gleaned from stakeholders, domain-specific knowledge, knowledge of rules and regulations, knowledge gained from experience on other projects, etc. Lack of a formal means of representing the knowledge shared among these agents often results in miscommunication which in turn results in poor decision-making and, thereby, schedule delays and cost overruns. Such issues can hinder the competitive advantage in a mission-critical environment. It is equally important to capture the knowledge possessed by systems engineers, who have a great deal of experience having worked on multiple long-term and large-scale complex projects, who are leaving the workforce. This paper focuses on formally capturing knowledge that exists in various phases of systems engineering lifecycle by leveraging epistemic modal logic. The approach in this paper aims to address some of the issues with the traditional document-centric approaches.

Our work is concerned with the development of model-based systems engineering (MBSE) procedures for multi-domain system architecting, configuration, and reasoning. This class of problems is characterized by the presence of multiple domains (which could be cyber, physical, or hybrid), each with their rules and constraints that need to be integrated in a correct-by-design systems and processes in order to satisfy stringent constraints on performance, safety, and effective management of system functionality. To that aim, there is a strong need for formal methods of analysis that can enable effective assembly of components into correct-by-design and provable systems capable of delivering desired capabilities. Thus, this paper discusses semantics and their central role in the development of a new ontology-driven knowledge modeling and reasoning approach for multi-domain system architecting and configuration. Three interdependent modules supporting each other are integrated to make up the system. A foundation module (1) with description logics (DL) as core knowledge representation formalism provides the necessary mathematical foundations to a semantic platform module (2) constituted of semantic blocks (i.e., ontology, rules, and specialized computational capabilities). A configurator module (3) later assembles systems from instantiated semantic blocks as per architectural configurations of interest to generate valid system design alternatives. An implementation of the system on rule-based generation of architectural configurations for satellite robotic arms demonstrates the capability of our approach.

The purpose of this paper is to investigate product development test processes. A literature review examines research on test activities in product design, product development and systems engineering research fields. The publications reviewed have been categorized based on the stage in development and placed into a proposed test process framework. The proposed framework sets an agenda of functions and characteristics important for the integration of test processes into model-based systems engineering. The findings presented are of interest to researchers by structuring test activities from product development, systems engineering and prototyping research into a context for the design process. The findings also allow practitioners to identify research at the level of planning and development stage relevant to their test processes.

The Modular Open Systems Approach (MOSA) is a DoD initiative that requires major defense acquisition programs to employ modular architectures using widely accepted standards. In order to realize the benefits of modular and open architectures, program stakeholders must successfully navigate various technical and programmatic decisions throughout the acquisition life cycle. Our observation is that many programs do not have sufficient methods and tools to perform analysis, assess trades, and produce evidence for decisions that produce good program outcomes in general and in specific respect to modularity. This paper presents a model-based approach to rigorously collect and present acquisition context data and data from analysis tools in a Decision Support Framework (DSF). Through an example multi-domain mission engineering problem, we demonstrate how the DSF enables comparison of modular/non-modular mission architectures in terms of cost and performance. In addition, an MBSE enterprise architecture model is used to implement the DSF and is shown to (1) provide detailed visualizations of alternative architecture solutions for better comparison; (2) allow traceability between features of the architecture and organizational requirements to better document adherence to MOSA principles; and (3) lay the groundwork for continued model-based engineering development downstream of the Analysis of Alternatives activity to the rest of the acquisition life cycle.

Changes are intrinsic to system development. These changes regularly cause cost and schedule overruns due to design rework. Existing techniques seem unable to provide an adequate change management process. Model-based systems engineering (MBSE) allows for the creation of new change management processes that improve on the traditional methods. A change management process for projects developed in a MBSE environment is outlined. This process depicts the advantages of using MBSE to reduce time spent on a project when faced with high-level system changes. This process was then applied in the development of a CubeSat project and reduced the time spent on changes compared to the traditional process.

Requirements in natural language, like shall statements, while commonly used, present inherent limitations in terms of accuracy and precision. Modeling requirements within a model-based systems engineering MBSE) framework shows promise to cope with these issues. Common approaches include either the definition of textual requirements as model objects or the flagging of system models as requirements. The first approach inherits the weaknesses of natural language. We show in this paper that the second approach necessarily leads to a poor set of requirements. We therefore argue that modeling languages, in particular SysML, need to be semantically extended to adequately model the problem space. We demonstrate with a specific example that simply flagging model elements as requirements is not effective to model the problem space. In fact, we show that such an approach produces a deficient definition of the problem space, since it inherently discards solutions that could otherwise be potentially acceptable to solve the problem that is being addressed. In addition, we leverage this example to discuss potential semantic extensions of SysML that could enable adequate modeling of the problem space that fulfills the formal conditions of good requirements.

Current methods of Model-Based Product Line Engineering (MBPLE) do not seamlessly extend to the management of variants of system simulation models. Both architectural exploration activities and product line verification and validation analyses could be streamlined by applying MBPLE to simulations. However, this new application requires careful consideration to mitigate new challenges associated with simulation variant management, such as consistency between architecture and simulation models, the addition of simulation context and model fidelity as new areas of variation, and management of both plant and controller models. This paper presents an in-depth literature review of previous work on variant management of simulation models, a discussion of the complexity of the problem, and a preliminary proposal for an approach of simulation model variant management in which SysML and variant modeling capabilities are used to define 150% black box representations of simulation models in order to automatically create full simulation models.

In this paper we offer an overview of concepts and implementation of the Executable Systems Modeling Language (ESysML). ESysML is a domain-specific language (DSL) developed in response to challenges regarding executability and support for time-based simulation models, which are often necessary for analysis in a Model-Based Systems Engineering (MBSE) effort. ESysML is loosely based on the Systems Modeling Language (SysML). It downsizes the language schema of SysML in favor of model execution. Consequently, only language concepts such as Block, Ports, Activity, Events, etc. which are essential for defining system structure and behavior are retained in ESysML. While this presents a tradeoff of language expressivity for execution, the objective here is to offer precise semantics for a kernel of SysML constructs which can be extended to support fit-for-purpose applications in other domains.

A SysML library and method for calculating system reliability and availability is described. The method can be used to model and predict reliability and availability early in the design and continue through detailed design and system integration. Values for failure rates, restoration rates, recovery probabilities, and other parameters are stored in specialized reliability blocks that are defined from a single parent reliability block, and equations for reliability block configurations (series, parallel, active/standby, and k out of n) have been implemented and included in the library. The paper includes an example demonstrating the application of the library model elements and how the library can be extended to include additional system redundancy configurations.

In the face of ever-increasing system and program complexity, several aerospace, automotive, and defense organizations have already begun transitioning to model-based systems engineering (MBSE). A key challenge that organizations face is determining whether it makes business sense and whether it is technically feasible to transition to MBSE given legacy and budgetary constraints. This paper presents a methodological framework for analyzing whether an organization is likely to benefit from MBSE implementation on large-scale system programs. In this approach, MBSE implementation is characterized in terms of: system complexity, environment complexity including regulatory constraints, and system lifespan. Twelve major industry sectors are evaluated to determine whether MBSE can be of benefit to that sector. Then cost-benefit analysis is used to justify the decision to invest in MBSE. The approach is generic and can be applied to different industry sectors.

The COSYSMO 3.0 systems engineering cost estimating model was introduced at CSER 2019 (Alstad 2019). That introduction summarized the sources for the model, notably COSYSMO 1.0 (Valerdi 2005), and explained its updated and new features; however, the actual impact of the new features was not analyzed. Specifically, these impacts were not discussed: the differential impact on larger programs of having Process Capability as a scale factor, rather than a cost driver; the quantitative results of the model’s solution to the impact-of-a-step-is-too-large problem; the possible numerical impact on size estimates due to COSYSMO 3.0’s greater emphasis on (exponential) scale factors versus (multiplicative) cost drivers; the comparative impact of the four size drivers; and an overall comparison of model features in COSYSMO 3.0 versus previous models.

Assurance cases are a means to argue about the safety, security, etc. of software systems in critical domains. In previous work, we presented a tool called MMINT-A to automate change impact assessment of assurance cases given system design changes. In this paper, we argue that applying model-driven techniques to assurance case development allows safety engineers and assessors to ask questions about these artifacts and answer them using automated tool support – something not achievable with traditional document-based approaches. To support this argument, we present a library of well-formedness constraints on assurance cases structured in the Goal Structuring Notation (GSN). The constraints are formalized using OCL and implemented in MMINT-A. We also discuss other types of constraint checks that are useful in the automotive domain given the ISO 26262 standard and internal company processes.

Model-Based Systems Engineering (MBSE) is expected to improve communication and consistency in system development over document-based approaches. As systems become more complex, modeling languages increase the information content of their semantics to simplify modeling construction and visualization. We hypothesize in this paper that such increase in the complexity of the semantics may be detrimental to the MBSE objectives of facilitating communication and consistency. We present the results of an initial survey in which we asked systems engineers individually to interpret the behavior captured by several models in the form of Systems Modeling Language (SysML) state machines. Significant discrepancies in their answers were found. In addition, we present a qualitative assessment of the potential implications of such interpretation discrepancies for system development, which include need for rework, modeling gaps, inefficient solutions, and solutions that are not fit for purpose.

Fuzzy logic provides a means to address uncertainty in selecting optimal elements in the systems design process. Using mathematical algorithms coupled with existing system engineering tools has the potential to increase stakeholder satisfaction and reduce lifecycle cost of the system. In this study we use Fuzzy TOPSIS algorithm to demonstrate realization of top-level design choices as well as enhancement of stakeholder satisfaction in the case of designing a prosthetic arm. Although prosthetic industry is taken as an example, this research is valuable in any situation where the exact intent of the stakeholder is uncertain; nevertheless, the systems engineer seeks an optimal product design that would meet stakeholder needs.

The functional decomposition allows description of complex system functionality with a hierarchy of simpler sub-functions. Determining the best set of sub-functions for any given function should be an orderly and methodical process. It is important that the method used captures the system without overlooking any necessary functionality. Of the many techniques, this paper will examine six (6) decomposition methods (operating modes, inputs and outputs, Hatley-Pirbhai template, processing rates, organizational structure, and matching the physical architecture). The LN-39, a standard inertial navigation system (INS), first deployed in the A-10 and the F-16, is used as a sufficiently complex exemplar for applying each method. Finally, the resulting LN-39 functional architectures are compared from a number of perspectives in the product development lifecycle to illustrate strengths and weaknesses of each decomposition technique. Each of the decomposition methods revealed aspects of the system function that were important to include in the functional architecture, and no single method was clearly the best. Finally, it is recommended that multiple methods be used to reveal a comprehensive set of elementary functions. Thereafter, it is recommended that these elementary functions are clustered and composed to develop the functional architecture.

We make the case that since model-based development of complex software-intensive systems has proven to be so effective, a model-based paradigm that encompasses assurance of the system makes excellent sense and will result in more rigorous, less ad hoc approaches to the development and maintenance of assurance cases. This will become especially clear in the manufacturing of autonomous motor vehicles. Adequate demonstration of the safety of autonomous vehicles is a huge challenge. Doing it once for a single vehicle is difficult. Doing it for multiple vehicles in a product family and coping with incremental changes in design from one model version to the next without redoing the complete safety analysis is even more difficult. We show that a comprehensive, rigorous model-driven approach to development and assurance holds the promise of more efficient and more effective assurance in general and also provides a mechanism for incremental assurance. We also briefly compare that with one of the current staples for documenting assurance cases – Goal Structuring Notation.

Advanced Driver Assistance Systems and higher-level automated features are rapidly being deployed in the automotive industry. A common development approach taken for ensuring safe operation of these vehicles is to focus on driving real vehicles in the planned operating environment. This approach has benefits, including helping to identify challenging driving situations a vehicle may encounter and providing evidence of safe operation. However, driving millions of miles during vehicle development does not scale as more features are deployed. A model-based engineering approach can be used to augment real-world driving to provide a more efficient method for developing safe features. This paper describes key elements of such a model-based approach that includes a mission plan containing key parameterized use cases that trace down to simulation scenarios used to identify scenario edge cases to support Safety of the Intended Functionality.

Autonomous systems incorporate varying degrees of adaptation behavior to sustain their operations with acceptable quality of service (QoS). The QoS capability of such highly complex dynamic adaptive systems depends on how well they respond to hostile external events. The paper formulates model-based assessment techniques to benchmark the QoS capability of a networked system of cars S. We elaborate on this approach with a MATLAB-SIMULINK-based case study of adaptive cruise control (ACC) system in automobiles: first, for in-vehicle CC and then for multi-vehicle coordinated ACC. We employ model-predictive intelligent control methods to dynamically adapt the ACC system configurations to attain optimal behavior.

Systems engineering process models provide the framework for visualizing and organizing a system life cycle guiding its design and development. The systems engineering Vee process model is ubiquitous for describing systems. It is the most widely used and often tailored for specific system types. This paper models the doolie cadet as a system of interest using the Vee model. The term doolie refers to a freshman at the United States Air Force Academy. The doolie year is a rigorous year with difficult requirements to complete the transformation of high school graduates into disciplined military academy cadets. This work describes the challenges associated with the doolie year and models the doolie cadet as a system of interest using the systems engineering Vee model. Beginning with decomposition along the left side of the Vee, requirements, subsystems, and components that need to be assessed for an evaluation of performance are identified, progressing up the right side of the Vee; this work details verification events such as exams, athletic assessments, and military tests and training. These events validate the system, demonstrating that the doolie cadet is proficient and ready for acceptance into the cadet wing. This work contributes a new perspective on modeling human performance as a system of interest and new insights into uses of the systems engineering Vee model.

There are two significant “players” in development projects: the project manager and the systems engineer. They work together with the aim of meeting the technical (execution/performance, quality) and managerial (schedule, costs, and customer satisfaction) goals of the project.The purposes of the current study (Kordova S, Katz E, Frank M, Syst Eng 22(3). https://doi.org/10.1002/sys.21474 ) are to identify the management processes shared by project managers and systems engineers in the defense industry; to understand which factors influence the ways in which joint project management is accomplished and how it impacts meeting project goals; and to provide recommendations for joint project management that will lead to project success.The research method was qualitative, based on 16 semi-structured interviews with project managers and systems engineers in defense companies that deal with the development of technological systems.The main recommendations for joint project management are: Set a clear distribution of responsibility and delegation of authority between the both parties before starting the project; choose a project manager who was once a systems engineer or who possesses knowledge of engineering; insist on ongoing dialogue between the two professionals; solve/prevent conflicts through discussion and persuasion; and expand the common ground between the project manager and systems engineer’s areas of responsibility.

Model curation is an explicit requirement under the first goal of the Digital Engineering Strategy. Model curation is the active archival of quality models. Model curation traces to digital and data curation; however, research in this area is within its infancy. A literature search identified the three major sources that are pioneering the field. This paper provides a highlighted overview of the three pioneering sources for model curation. The US Army Combat Capabilities Development Command CCDC) Armaments Center (AC) is prioritizing the transformation to digital engineering through Model-Based Systems Engineering and the Integrated Model-Based Engineering Environment and has begun to investigate how to implement a model curation capability.

As systems grow in complexity, systems engineers have embraced Model-Based Systems Engineering (MBSE) to tackle this complexity. The Systems Modeling Language (SysML) is the most commonly used language by the systems engineers to implement MBSE. SysML is not highly capable of expressing conceptual but not executable models. In order to perform requirements/behavior verifications, systems engineers/designers mostly use separate simulation tools. Hence, the efficiency of the systems engineering process is often reduced due to the isolated and consecutive use of both SysML modeling tool and other simulation tools, for example, defining simulation inputs to each simulation tool separately. Hence, executable SysML is the next logical step towards achieving true MBSE support for all systems engineering activities in the life cycle phases – system requirements, analysis, design, implementation, integration, verification, transition, validation, acceptance testing, training, and maintenance. Therefore, various research efforts are being conducted to develop executable SysML modeling approaches. This research develops a SysML Executable Modeling Methodology (SEMM), which is demonstrated by modeling a CubeSat-based Space Situational Awareness (SSA) system in SysML. The SysML SSA-CubeSat system model is made executable by integrating with Commercial-Off-The-Shelf (COTS) simulation software, namely, Systems Tool Kit (STK) and MATLAB, following the approaches defined in the SEMM.

Manufacturability involves many different influence factors from product design and geometry to supply chain and ergonomics. Many software packages are available for product assessments; however a gap in the available software packages was discovered. The lack of a software that looked at both product design and the design of the process, along with other variables that affect the total manufacturability, was identified. From this research, the MAKE tool was created; however how to score and weigh each assessed part of a product was still under investigation. This paper outlines a comparison between two proposed weighting methods for a manufacturability assessment. The first was the topic of a 2019 CSER paper describing a weighting method conducted by SME inputs and counting risk concerns and high scores. The second method, which is introduced in this paper, uses a value curve method to assign weights to each aspect of manufacturability. A case study is used to illustrate each method, and the results are compared and graphically displayed to magnify similarities and differences between methods along with conclusions and future research areas.

The objective of the proposed research involves the challenge of developing a methodology and tool to assess the manufacturability of conceptual designs at Milestone A, where minimal system design information is available. From a practical standpoint, the idea of utilizing a subject matter expert SME) as a basis for judgment on a design’s manufacturability early in the design process lacks feasibility due to the inability to efficiently and effectively evaluate a large tradespace of unique design alternatives. The practice of Design for Manufacturing (DFM) analysis typically involves having access to design geometry and specifications with some consideration of the manufacturing processes. However, in the conceptual stage, the challenge involves assessing manufacturability based on a significant number of unknown parameters and doing so in a manner which is nonsubjective. Furthermore, evaluation of early stage product designs has significant influence on program cost. So, how can programs realize the impact of design options that may influence manufacturability and relate this back to a common frame of reference (i.e., cost, schedule, risk)? A research challenge includes determining how to harness the knowledge that is used to determine manufacturability from both factual and heuristic-based approaches, which requires some knowledge of the design parameters and the decision-making involved with assessing manufacturability. There are different ways to explore this area of research, but one possible approach rests in the exploration of artificial intelligence and how it can be applied in the area of manufacturability assessments. There are various subsets of artificial intelligence; some involve areas such as rule-based engines and systems, knowledge graphs, and expert systems, while others explore more complex areas such as machine learning and neural networks. The choice on which path to take requires some exploration into these possibilities and an understanding of the design data available in pre-milestone A and how feature-based information can be used to create an objective-based manufacturability assessment. This paper serves to explore the options for incorporating artificial intelligence within the MAKE assessment methodology and related software tool. Based upon feedback from the user community, one or more of these options could be incorporated in future efforts.

This research investigates a bioinspired framework for analyzing and predicting trade-offs between system of systems’ (SoS) performance, affordability, and resilience early in the design process – without the need for highly detailed simulations or disruption models. This framework builds on ecological research that has found a unique balance between redundancy and efficiency in biological ecosystems. This balance implies that highly efficient ecosystems tend to be inflexible and vulnerable to perturbations, while highly redundant ecosystems fail to utilize resources effectively for survival. Twenty architectures for a notional hostiles’ surveillance SoS are investigated, showing that highly efficient SoS architectures fail catastrophically in the face of disruptions, while highly redundant architectures are unnecessarily expensive: indicating that engineered SoS architectures follow a fitness trend akin to complex ecological networks. The results suggest that SoS may benefit from mimicking a balance of redundancy and efficiency similar to that found in ecological networks.

Achieving value-based healthcare – increasing quality, reducing cost, and spreading access – has proven to be extremely challenging. In recent years, a large variety of care coordination organizations have emerged at regional and national scales. Unfortunately, each such health entity lives in its own definition (silo) of care coordination leaving large gaps in care as well as duplicative or inconsistent interventions where care domains overlap. This situation leads to the need for higher-level coordination of the coordinators with well-defined population health metrics and means for sharing of information and control of patient-centered interventions. In “Value-based Learning Healthcare Systems: Integrative modeling and simulation” (Zeigler et al. 2018), we presented a modeling and simulation (M&S) approach to value-based healthcare within a system-of-systems framework that enables designing, testing, and implementing care coordination based on identifying and addressing risks at the individual and family level and tracking progress though health information technologies (HITs). In this paper, we discuss how a model-based system-of-systems design for HIT infrastructure can support innovative “coordination of the coordinators” assuring that critical modifiable risks spanning health and social issues are identified and addressed resulting in better health and social outcomes. We describe existing foundations for implementing such a design such as digital platforms, pathways-based community coordinator organizations, risk factor registries, as well as comprehensive simulation facilities where the design and its components can be tested. Research required to enable integrating such foundations into a working whole is also described.

The CubeSat standard has given universities, small companies, developing countries, and others a new gateway to space exploration and space knowledge. Combined with shorter development time and Commercial Off-The-Shelf components, the cost has been lowered considerably. However, the combination of use of low-maturity components and inexperienced development teams results in a short lifetime and poor reliability for most CubeSats. The growth of model-based systems engineering (MBSE) supports reuse of design architectures in many industries and has lowered the costs of development and is gaining popularity in CubeSat teams. This paper demonstrates the application of reliability methods for implementing dependability analysis in MBSE and shows how this can benefit CubeSat teams struggling with limited personnel resources and low experience with space systems.

As technological advances in material science, computing, electronic, and artificial intelligence are fueling increasingly autonomous unmanned aircraft system (UAS) capabilities, the legal operational framework is progressively catching up. However, for commercial applications involving beyond-visual-line-of-sight (BVLOS) group flights to become ubiquitous, the successful development of safe, integrated UAS traffic management (UTM) systems (UTMSs) is imperative. We propose a generic model-based systems engineering (MBSE) approach for supporting conceptual design and analysis of UTMS. A system-of-systems (SoS) perspective of the UTMS is adopted to drive the design effort. This iterative process is rigorous and technology independent and is powered by the flow-down of requirements driving the specification of each layer of abstraction of the UTMS as a SoS. At each layer (SoS, systems, subsystems, components), functional, structure, and behavior are developed and synthesized and then verified against the requirements. Various analyses, including trade studies, are performed to support system evaluation and guide decision-making throughout the design. The essential features of the framework are highlighted in a simplified UTMS as a directed SoS for a package delivery application developed using the Systems Modeling Language (SysML).

Design reuse, a common and valuable practice in complex engineering projects, is particularly important within the space industry. However, design reuse in this discipline is often conducted in an ad hoc fashion that limits the upside while introducing potential pitfalls. Model-based systems engineering (MBSE), while in the early stages of adoption in the industry, may form the foundation of a new strategic and systematic methodology for investigating and implementing design reuse for space missions and campaigns. To that end, this paper explores one aspect of that envisioned methodology, namely, the technical inheritance process which assesses the technical feasibility of adapting design elements from their native missions to new missions of interest. First, we investigate the impact of the level of architectural decomposition of the element to be reused. Then, we synthesize a generalized version of the process from a historical survey of design reuse examples and suggest MBSE methods, specifically using SysML, to implement them. Lastly, we discuss where this inheritance process fits in the larger picture of the systematic reuse strategy.

Models in systems engineering have existed in various forms dating back to the 1950s. They have been used by engineers to understand various types of phenomena, envision future systems, and generate engineering artifacts. Today the increasing complexity of operational missions and technological advances enabled in part by disciplinary convergence and wide access to data are having a dramatic impact on system modeling. Operational missions are becoming increasingly more complex with multiple sources of uncertainty and subject to a variety of disruptions. Technological advances paced by advances in semantic technologies, machine learning, AI, and applied analytics are transforming model development into a closed-loop process. The advent of Industry 4.0 and digital engineering (including digital twin and digital thread) is causing models to be viewed in an entirely new light. And the convergence of engineering with other disciplines is opening up a whole new way of developing system models. This paper presents a historical perspective on models over several decades and offers a vision of how recent developments are likely to shape the trajectory of system models in the future.

Transdisciplinary systems engineering provides new ways of thinking when developing engineering solutions to complex systems problems. The solution space afforded through the use of transdisciplinary approach tends to be broader than that achievable with traditional approaches. This expanded solution space is a result of new ways of thinking enabled by exploiting disciplinary convergence. A shift to transdisciplinary systems engineering requires a shift in the approaches and tools needed to engineer complex systems. These changes in systems approaches include models that can capture transdisciplinary concepts, inclusion of expansive meta-data in transdisciplinary models, and application of nontraditional engineering methods to explore concepts and outcomes. As engineering takes on the challenges of increasingly more complex systems with different levels of autonomy, we believe that transdisciplinary approaches will enable the realization of more complete solutions for such systems and enable the complexity necessary for these systems. This paper presents specific approaches within the rubric of transdisciplinary systems engineering.

Compositionality reasoning is fundamental to engineering. The problem of compositionality is typically framed as: given a configuration of parts with characteristics and interrelationships, how can we derive the characteristics of the configuration as a whole? This paper uses systems science concepts to address a related scope question: how should parts be characterized, and what are the kinds of compositionality reasoning needed, in order to assert that a given configuration will exhibit desired behaviour and characteristics?Our model structures compositionality reasoning into categories that take a successively wider view of the system: pattern of organization, variety, dynamics, planes of operation, levels of organization and context impacts. The ultimate objective is to explicate the systems science basis for systems engineering and contribute to tooling and engineering practice by identifying the compositionality assertions to be satisfied and the scope of concerns to be covered in systems modelling. To test the coverage and utility of the proposed model, we examine its relationship to systems engineering practices in a large radio telescope project and the extent to which it covers a collection of systems science concepts identified in the literature.

Many complex systems evolve as a result of interactions among individual entities whose behaviors cannot be directly controlled. This makes the design of such systems inherently challenging. The objective of this research is to develop a new approach in engineering complex swarm systems with desired characteristics based on the theory of network motifs – subgraphs that repeat themselves among various networks. In recent studies, the discovery of network motifs has presented the ability to determine reoccurring similarities between similar functioning networks that were originally believed to have not shared any characteristics. It is therefore hypothesized that manipulating the types of network motifs within a network can help engineer artificial swarms with improved functionality. In this study, artificial swarm systems have been modeled as a dynamic complex network where each node represents an individual foraging entity and links represent as the communication between entities. Additionally, motif-detecting algorithms have been used to extract subgraphs that reoccur in these complex networks. Our research has shown promising results that reveal a statistically significant correlation between network motifs and the performance of simulated swarm networks. This study contributes as a new approach that can potentially be used in the design and engineering of complex swarm systems.

Integrating small unmanned aerial systems (sUAS) into the National Airspace System (NAS) represents a challenging problem set that requires consideration through multiple lenses. Rather than focusing solely on the technological limitations of sUAS operation, this work employs the Architecting Innovative Enterprise Strategy (ARIES) Framework to understand the current and future landscapes for the NAS. The authors use the ARIES elements to holistically describe the current architecture that allows for limited sUAS operations in low-altitude urban airspace. The goal of this work is to develop a level 1 CONOPS as a first step toward a complete enterprise architecture. By identifying current limitations and incorporating emerging concepts and technologies, the authors develop a realistic envisioned future. This future seeks to address externalities that emerge from the increased use of sUAS near the general public. Specific externalities include safety, security, privacy, and transparency concerns. The envisioned future relies on airborne systems to detect and avoid manned aircraft and utilizes an unmanned traffic management system for information sharing and flight coordination. It requires significant investment in developing a shared database to manage unmanned vehicle operations while providing the structure and functions required to make sUAS operations feasible, considering constraints, externalities, and public acceptance.

The lack of and need for theoretical foundations to systems engineering have been recognized by multiple researchers in recent years. The lack of a foundation extends to the positions of systems engineers in organizations. Presently, organizational architectures for systems engineers are based on heuristics. Since systems engineering is required to alleviate certain challenges associated with the development of complex systems, a strong theoretical foundation to the establishment of organizational architectures for systems engineers is imperative. Such a theoretical foundation will ensure that the contribution made by systems engineers to organizational value can be improved. The goal of this paper is to provide a basis for creating a mathematical framework for organizational architectures for systems engineers. A literature review spanning multiple disciplines is conducted to identify elements pertaining to the organizational architectures. These elements are then used in a directed graph to visually represent the relationships between the elements and the mapping between systems engineers and organizational value.

This paper presents a proposal for the application and modelling of systems engineering to an already deployed system as a means of resolving challenges frequently associated with deployed enterprise content management systems. It generalizes traditional systems engineering methods as a model that may be applied to already deployed ECM systems. In this paper, the author reviews the traditional systems engineering process followed by a more detailed review of the modified, post-deployment systems engineering process. The author then concludes by discussing the benefits of performing post-deployment systems engineering.

The digital transformation era is upon us. Digital transformation gradually crawls up the value chain from services and manufacturing to product design and systems engineering. In this paper, we envision a cloud-based ecosystem of systems engineering, which is model-based by definition. The ecosystem model we propose is called 2MIDSTARs, which stands for Model, Infrastructure, Data Services, Simulation, Testing, Analysis, and Repositories + Management, Interoperability, Digital Representation, System, Technology, Audit, and Reporting. The first MIDSTAR covers the intrinsic, core MBSE capabilities, while the second MIDSTAR facilitates the integration with the digital enterprise that surrounds the digital systems engineering ecosystem. In this paper, we explain the importance of jointly considering all these elements together and outline the key roles and functionalities of each component.

Engineering projects that implement model-based systems engineering encounter the issue of connecting the managing functions of a project with the system engineering model. This research uses a pilot case study to demonstrate how the management of systems engineering can be done in SysML models within the Open Model-Based Engineering Environment (OpenMBEE). The management ability for this pilot includes continuous updates, web-based collaboration, model-based report generation, and enabled semantic reasoning. The semantic reasoning is seen as a key enabler and accomplished using a SysML profile that is aligned to an underlying project ontology. This not only results in utilizing the advantages of a model-based engineering environment for managing the project but also demonstrates the benefit of semantic enabled reasoning that is a focus of research of the research project.

Dynamic Risk Analysis (DRA) approaches are virtuous processes enabling the improvement of state-of-the-art techniques for risk calculation in industrial infrastructures. However, they require the existence of an appropriate architecture enabling end-to-end processing of information, which has not yet been defined in practice. This paper aims at discussing the possibilities and the advantages of combining DRA with Systems Engineering (SE) approaches to reach this objective. For that, we define a framework based on SE principles, apply it for the assessment of the role of vegetation on the global risk for power grids, and discuss the benefits it provides.

The historical failure data reliability prediction method commonly used by systems engineering practitioners has several limitations. Recent literature promotes the physics of failure reliability prediction approach and has seen limited adoption. However, there is limited guidance available to practitioners to determine when the historical failure data reliability approach is appropriate to use and when the physics of failure approach is best applied. This paper presents a decision support framework for practitioners to choose between historical failure data and physics of failure reliability approaches and is specifically meant to be used in early system design. The Reliability Decision Framework RDF) identifies key factors in system design that aid practitioners in the selection of an appropriate reliability prediction approach for systems of interest. The major factors in the decision are (1) relevant historical data, (2) level of complexity, (3) operational life requirement, and (4) criticality of the system.

The annual threat of increasingly severe Atlantic hurricanes has raised concerns about the management of logistics in the immediate aftermath and rebuilding phases of catastrophic storms. Logistics challenges include timely delivery of inbound relief supplies, synchronization of supply and demand in the reconstruction phase, and management of the reverse flows of empty containers. Due to the low probability of occurrence, the high level of impact, and the scale and complexity involved, it is often difficult to devise a comprehensive logistics plan in advance of a catastrophe. To ensure satisfactory performance at all stages of recovery, the problem must be conceptualized in systemic terms using information from past experiences to identify causal relationships and opportunities for optimization. This paper highlights the challenges faced at Caribbean ports in the immediate aftermath of a hurricane disaster and in the reconstruction materials supply chain during the rebuilding period and offers a conceptual approach to improve planning by combining holistic thinking with simulation and information technologies.