MODEL-BASED PRACTICES
Model-Based Practices
The Future of Systems Engineering Is Predominantly Model-Based
TO
Systems engineers routinely compose task-specific virtual models using ontologically linked, digital twin-based model-assets. These connected models are updated in real-time providing a virtual reality-based, immersive design and exploration space. This virtual global collaboration space is cloud- based, enabled by modelling as a service and supports massive simulation leveraging cloud-based high-capacity compute infrastructure. Families of unified ModSim frameworks exist enabling small and medium businesses along with Government agencies to collaborate.
FROM
Although a growing number of systems engineering organizations have adopted model-based techniques to capture systems engineering work products, the adoption is uneven across industry sectors and within organizations. Custom, one-off simulations are used for each project, and there is still limited reuse of models especially during critical early phases of systems architecting and design validation.
By 2035, a family of unified, integrated MBSE-Systems Modeling and Simulation (SMS) frameworks exist. They leverage digital twins and are fully integrated into the enterprise digital thread foundation. This enables efficient pattern-based model composition and seamless “cradle to grave” virtual exploration. Integrated AI/ML-based agents identify high impact parametric studies, noise factor sweeps, and support closed loop safety/security operational domain design surface explorations.
The digital thread-based MBSE-ModSim framework(s) enables agile and efficient capture, modeling, simulation, and understanding of user experiences. Virtual design candidates can be evaluated down through manufacturing, maintenance, updates, and eventual decommissioning. The MBSE-SMS frameworks also leverage high bandwidth, bi-directional connectivity supporting fresh data ingestion, segmentation and AI/ML network re-training. Real-time system anomaly detection can be detected and the connected data used for virtual systems design updates. These updates can be deployed to the field using exploratory “shadow software” to run in parallel to the main software application, gathering performance data and informing the systems engineer on appropriate release actions. Finally, the MBSE-SMS frameworks provide integrated asset life cycle management systems that support Agile continuous integration, build, validation and release cycles.
MBSE-SMS FRAMEWORK
CONNECTED DATA
Highly connected data with integrated AI/ML-based data segmentation, object labelling, and temporal scenario – ontology mapping supports automated digital twin creation, model correlation, verification and validation and seamless systems engineering trade studies.
MODEL-BASED SYSTEMS ENGINEERING
MBSE Descriptive models created using semantically rich modeling standards provide systems abstraction, data traceability, separation of views, and leverage AI/ML-based reference model reuse at both systems and product realization levels.
INTERACTIVE HMI VIRTUALIZATION
Interactive customer HMI experiences with virtualized connected services, real-time control algorithm, and CPU emulation providing real-time system response parameter exploration.
Increasing Fidelity and Completeness Supporting Extended Reality
Layered simulation models at multiple levels of abstraction allowing real-time simulation at multiple scales (single vehicle to multi agent traffic to city infrastructure fleet to regional/cross country simulations)
DETAILED SCENARIO ANALYSIS WITH PHOTO-REALISTIC VISUALIZATION
Photo-realistic simulation and visualization enable detailed scenario analysis.
SYSTEMS OF SYSTEMS
Real-world Systems of Systems, operational design domain customer experience data into cloud-data-lakes providing instantaneous opportunities for action-oriented information extraction.
GAMING ENGINE PHOTO REALISM AND EXPLORATION
Extended Realities: (xR) Augmented, Virtual, and Mixed.
MASSIVE PARALLEL COMPUTE
High-Capacity Parallel Compute supporting advanced AI/ML augmented data visualization providing synthetic data generation, and deep learning-based edge case exploration for performance, safety-risk, and security-threats.
ENVIRONMENTAL CONDITIONS, TOPOGRAPHIES, SCENE GENERATION, AND MAPS
High fidelity 3D maps, road topologies, scenes, weather and traffic conditions.
Architecting Flexible and Resilient Systems
TO
Systems architecting methods are well established and address broad stakeholder concerns associated with increasingly complex systems. Systems architecture, design, and analysis is integrated across disciplines, domains, and life cycle phases to provide a single, consistent, unambiguous, system representation. Evolving needs, external context, and anticipated failure scenarios are central to the architecture process resulting in flexible, resilient, and adaptable architectures.
FROM
Systems architecting is often ad-hoc and does not effectively integrate architectural concerns from all relevant technical disciplines (such as hardware, software, operations, manufacturing, security, and more) nor does it fully integrate other stakeholder concerns. Further, systems architecting does not always include sufficient environmental context or failure scenarios to evaluate and optimize the architecture for operational resiliency.
Engineered systems have always been used in ways that were not considered during their initial design, sometimes adapting elegantly to new use cases. However, as we approach 2035, designing systems and their supporting systems and supply chains with specific focus on flexibility, robustness, and resilience will be a central tenet of the architecture process. Emerging techniques such as Loss-Driven Systems Engineering (LDSE) and Opportunity-Driven Systems Engineering (ODSE) will help systems engineers identify systems optimizations to increase systems resiliency. Techniques such as chaos engineering will be adapted to drive resiliency of a greater variety of system types (not just IT and software systems).
Resilience pursues a future where systems have the ability to deliver required capability in the face of adversity. Systems engineering practices by 2035 will design systems that can adapt to emergent systems and operations behaviors in both reactive and proactive ways.
The emergence and commoditization of autonomous systems illustrates the need for systems resilience as these systems must be robust to a wide range of environmental conditions, adaptive to unexpected conditions, and capable of anticipating and recovering from failure conditions. Resilient systems can continue to carry out the mission in the face of disruption, and by 2035, systems engineers will readily use high fidelity modeling, simulation, and analysis to evaluate and optimize systems to be resilient to various operating conditions, failure scenarios, and unexpected conditions.
Resilience Architectures in Smart Cities
Smart cities integrate data from a vast array of sources–deployed sensors, buildings, transportation systems, utilities, and more. This data is used to both inform decision makers, but also to automatically react to changing conditions. The highly-interconnected nature of smart cities and the potential for interdependence between municipal functions drives a need for a highly resilient architecture.
A resilient smart city architecture will address and limit the risk of cascading intra and inter system failures, support integration across systems, and will facilitate continuous, dynamic adaptation, and expansion of systems of systems
Engineering Trusted Systems
TO
Systems engineering routinely incorporates a range of new perspectives including security, privacy, and explainability with traditional perspectives such as systems safety to define and track a metric of “systems trust”. This includes designing with data minimization and defense in depth principles to protect the systems from cyber- threats and minimize the impact to users if a system is breached.
As autonomous systems become mainstream, principles of explainability and provable safety will allow system providers to build confidence in these systems and will allow those system developers to differentiate themselves in the marketplace.
FROM
Systems trust is a loosely defined concept that includes many properties including cyber-security, data privacy, systems safety and overall reputation. The legal landscape governing how systems must address these properties is evolving quickly and inconsistently, but the properties that comprise “trust” are routinely “secondary” considerations in overall system designs. But the increasing level of interconnectedness in systems and the increasingly routine nature of data collection to power new systems, is resulting in a risk surface for organizations that is rising exponentially.
Transparency and Corporate Ethics
System properties only make up one portion of the trust equation; system developer behavior and country of origin also contribute to how users feel about systems. By 2035 corporate ethics, reputation and transparency – especially regarding use of personal data will be central to how users determine what systems to trust, and which to avoid.
CYBER-SECURITY
The cyber landscape is ever evolving with new threats emerging daily, including a wider variety of nation-state actors forming attacks for political, strategic, and economic gain. As our digital infrastructure becomes increasingly connected and we begin to rely more heavily on autonomy, cyber- security is increasingly a major tenet of systems safety and forms a foundation of trust.
By 2035, cyber-security will be as foundational a perspective in systems design as system performance and safety are today. The systems engineering discipline will grow to become even more interdisciplinary, embedding cyber expertise into the team to ensure cyber is considered through the full system life cycle. Additionally, modeling and simulation tools to help test and evaluate cyber aspects of the system will be increasingly prevalent, providing a holistic picture of system security that is too often only considered late in the development life cycle today.
Design for cyber-security will extend beyond the components of the system to include analysis of the supply chain and sourced parts to eliminate any weak spots in the system.
DATA AND PERSONAL PRIVACY
Systems are increasingly reliant on collected data to operate. Data is critical to the functionality of autonomous systems, and other systems that learn and adapt to user preferences and behaviors. Users will increasingly trust system providers that are responsible with user data, transparent, and have mechanisms for data minimization and protection surrounding any and all collected data that is personal in nature.
Infusing Data Science Methods into Systems Engineering Practice to Understand Complex Systems Behavior
TO
The systems engineer’s toolkit is significantly expanded by the infusion of data science tools and techniques, allowing better quantification of performance and risk for non-deterministic systems, and improved ability to continually monitor system behavior over time. This expanded toolkit facilitates more comprehensive analysis and allows for the collection and exploration of extremely large, interconnected data sets to understand increasingly complex systems.
FROM
Systems engineers analyze system behaviors using models of performance, physical constraints, cost, and risk using a mix of tools ranging from commercial simulation and analysis tools to spreadsheets to homegrown code. These analyses generally are restricted to relatively limited data sets and systems engineering practice does not generally include methods to correlate large data sets to help understand complex behavior.
The increasing complexity of systems of 2035 also increases the difficulty in analyzing and predicting systems behavior. Cyber-physical systems will be massively interconnected, incorporate smart systems technology, and must be safe and trusted. Systems engineers will be expected to analyze these systems with increasingly large trade spaces and extremely large data sets to quantify system behaviors. Systems engineering practices will require smart data collection mechanisms and will include both formal and semi-formal methods for identifying emergent behaviors and detecting, quantifying, and managing uncertainties and unanticipated behaviors leveraging that data.
Improvements in data science methods and open-source tools coupled with inexpensive cloud-computing resources will help power the next generation of systems engineering practices and tools, allowing the systems engineer to better understand possible non-deterministic outcomes while also coping with uncertainty. Research in data science, data analytics, and big data will be infused into the systems engineering practice and data science will become a core competency of the systems engineer
Analytical techniques adopted from the data science discipline such as clustering, outlier detection, and probabilistic reasoning, will be commonly used to explore huge systems state spaces to identify and eliminate undesirable systems states. Techniques will be developed to correlate, monitor, and visualize a diverse range of systems parameters as indicators of systems health. Analytical techniques will leverage large data sets from real-time monitoring of operational systems, that is used to better understand the systems behavior and improve systems performance and other quality characteristics. Capitalizing on this understanding to develop systems that are more fail safe, fault tolerant, secure, robust, resilient, and adaptable will be a fundamental part of systems engineering practices. Visualization tools will enable interactive analysis from many different stakeholder-specific viewpoints, allowing decision makers to gain new insights, perform what-if analyses, and communicate the impact of their decisions.
Systems engineers and decision makers will have more information and machine-driven insights from which to draw conclusions
Model-Based Systems of Systems Practices
TO
Systems of systems are designed with a family of unified modeling approaches. Common SoS style guides, patterns, and methodologies are practiced that integrate socio-technical systems, human factors, cybernetics, evolutionary biology, and sociology into the SoS. Model-based verification of SoS’s are realized using service-agent based model composition on a stand-alone or net-centric simulation platform. Systems engineers design-in assumed SoS reuse within an anticipated larger SoS solution.
FROM
The systems engineer is primarily focused on the design of dedicated domain specific systems. There is broad recognition that systems and devices are no longer stand alone but are interconnected as part of broader systems of systems (SoS). Initial design guidance has been developed in the form of architecture frameworks and interoperability standards.
By 2035 the systems of systems engineering (SoSE) community has grown to include practitioners across a diverse set of domains including Government-Policy, Civil and Commercial.
These communities have identified the collective advantage of working together and treating the aggregate set of separately owned and operated technical and non-technical systems, and applying a broad-based systems approach despite the lack of a ‘top level authority’. This opens new opportunities for implementing SoSE across domains.
SoSE has evolved to include aspects of Socio-Technical Systems Theory, Open Systems Principles, Network & Network Analysis, and Interoperability Models into the systems engineering best practices.
Collectively, these practices provide the SoSE with a core set of frameworks to capture and analyze SoS in terms of legal, organizational, semantic, and technical interoperability. These SoS frameworks also have gone a long way to address the key challenges identified in the INCOSE handbook.
New SoSE patterns have been established that are leveraged to design and implement extensible, robust and adaptive SoS solutions. These patterns include object oriented systems engineering (OOSE) methods such as data encapsulation, inheritance, and abstraction. These model-based techniques fully integrate SoSE-patterns, OOA/D and AI/ML network analysis providing an extended capability to explore the full virtual SoS concept space. They are used to design an extensible and re-usable systems in the context of systems of systems.
“A SoS is an integration of a finite number of constituent systems which are independent and operable, and which are networked together for a period of time to achieve a certain higher goal.”
Understanding Socio-technical Complex Systems with Human Systems Integration Methods
FROM
While there is a notable increase in the adoption of user experience design methods, there is still a gap between systems engineering and user experience teams. Systems analyses often focus only on the technology-centric aspects of systems or model the human in the systems with limited fidelity.
TO
Systems engineering methods have integrated User Experience concepts and Human Systems Integration (HSI) methods to ensure human elements of systems are appropriately understood and considered in all aspects of design. Further, as systems more frequently rely on human- machine collaboration, the interplay of the technical elements and human elements of the system will become an increasingly critical part of system design.
Approaching 2035, socio-technical systems will be increasingly autonomous, incorporating more AI, will be massively interconnected, and must be collaborative, safe, secure and trusted. Analyzing and predicting system behavior will become more challenging, but systems engineers will be expected to analyze, design, and evaluate these systems with human and natural principles in mind. Systems engineering practice will include HSI methods for evaluating human factors and usability, identifying emergent behaviors, and detecting and managing unanticipated behaviors.
Improvements in HSI methods, human behavioral simulation and human-in-the-loop simulation capabilities will help power the next generation of systems engineering practices and tools, allowing the systems engineer to better understand possible non-deterministic outcomes and cope with uncertainty.
Research in HSI will be infused into the systems engineering practice and become a core competency of the systems engineer.
The human will be increasingly “part of the system” to solve complex problems and will have a wide range of interaction mechanisms available ranging from voice to touch with haptic feedback.
By 2035, human-machine interfaces have continued to evolve, following current trends, providing users with a wide variety of ways to interact with systems, including voice, touch, and gesture. HSI will increasingly focus on human-machine collaboration as more humans, machines, and processes to solve previously intractable problems.
HSI generally incorporates various dimensions that need to be integrated: human and organizational factors, HSI planning and project management, manpower and evolution of jobs, personnel, training, life-criticality that includes occupational health, safety, environment, habitability, and human survivability. HSI is interested in socio-technical complex systems with respect to systems of systems topology, human and machine activities and emergent properties. Systems interact among each other through various kinds of organizations, communities and informal groups. HSI includes the perspective of all personnel ranging from system owners to operators, maintainers, support personnel and end users.
Shifts in Acquisition Towards Collaborative Processes
FROM
Project needs and requirements are prepared ‘in-house’ by organizations to inform traditional acquisition processes, with the consequence that the project does not fully leverage the knowledge of the wider enterprise during its earliest and most formative phases. Acquirers possess limited ability to assess technical performance during the systems development process, while contracted parties are not motivated to share information. Reference architectures, when used, are unique to projects and not maintained after delivery of the systems.
TO
Acquiring organizations leverage industry knowledge during the earliest phases of a project, prior to the ‘main contract’. They establish multi-organization integrated project teams to perform as ‘smart’ customers during the entire systems life cycle, able to build upon evolving reference architectures and best practices. Shared digital engineering solutions maximize access to, and enhance the use of, information by all project participants during all phases, including ‘smart operations’.