Rigorous requirements elicitation, development, and validation form the foundation of systems development. A system built precisely to the wrong specification still fails. Ambiguous stakeholder needs are translated into precise, testable, human-centered specifications — elicitation workshops that surface latent needs, formal specification of functional constraints, traceability linking goals to system behavior, and risk analysis focused on human error. Context of use analysis, Operational Design Domain specification, and Operating Envelope definition ensure that what is specified can be built, integrated, and operated by the human teams who will work within it.
ODD · OES · Traceability · Risk AnalysisIntegration of human-AI teaming frameworks into system design, development, commissioning, and operations — allocating tasks to the agent best suited to perform them at each level of abstraction, and determining the appropriate level of automation for each allocated task. We cultivate intelligent systems. We do not merely design them.
Task Allocation · Levels of Automation · LifecycleAI behaves differently than traditional software, and standard quality assurance is not sufficient. Evaluation must examine how human and AI agents actually interact under real operational stress — red-teaming for prompt injection and behavioral drift, usability testing for explainability and calibrated trust, human-in-the-loop workflow evaluation, and bias detection in decision-support systems. Measurement spans human, AI, and organizational agents, aligned to TRL, IRL, and HRL readiness frameworks from the aerospace, defense, and automotive industries.
Red-Teaming · Explainability · Bias DetectionStructured co-design across the full human-AI-systems ecosystem. Intelligent adaptive systems cannot be fully specified before deployment — the social component must be incrementally evolved. Participatory design ensures that the humans who operate a system co-develop the tools and procedures they will use, adapt them over time, and cultivate resilient, variety-increasing teams.
STS · Participatory Design · Co-DesignMethods & Frameworks
Ensuring that human-AI system assessments, models, and simulations reflect real-world conditions — context of use, operational design domain, and multi-sensory, real-time data — so that systems that perform in the lab also perform in the field. Grounded in Klein's naturalistic decision making research and applied to human-AI team development and validation.
Ecological Validity · NDM · Context of UseSimultaneous tracking and alignment of Technology, Integration, and Human Readiness Levels across program development phases — ensuring that as systems mature technically, the human and organizational components mature in parallel. Based on ANSI/HFES 400-2021 and the HAISTE (ANSI/HFES 900) standard framework for human-AI system test and evaluation.
TRL · IRL · HRL · HAISTEAssessment and design of situation awareness at both the individual and team level. For individual agents: perception, comprehension, and projection of system state across human and AI agents in complex, dynamic environments, grounded in Endsley's SA model. For teams: shared mental models and common operating pictures ensuring that all agents in a joint cognitive system maintain compatible representations of system state, goals, and available actions — essential for effective task allocation and closed-loop operation in systems with extensive automation.
SA · Shared SA · Joint Cognitive SystemsSystematic analysis of the cognitive and physical work performed by agents in complex sociotechnical systems — identifying functional purposes, abstract functions, generalized functions, physical functions, and physical forms across the abstraction hierarchy. Provides the analytical foundation for requirements engineering, interface design, and task allocation in AI-enabled systems.
CWA · Abstraction Hierarchy · Work Domain AnalysisDesign as ethical act
Design of the Context of Use can enhance or detract from agents' physical and cognitive performance, their social interactions, and their well-being. Every design decision is, at its root, an ethical decision — whether or not it is recognized as such by the designer.
Ethics, as it applies to the design of intelligent adaptive systems, operates at two levels. At the first level: whether use of a particular technology — or how it is implemented — is detrimental or beneficial for the agents who use it. At the higher level: how systems of coevolving agents and technologies achieve collective benefit through mutually beneficial symbiotic relationships. The wellbeing of any agent in a system is not independent of the wellbeing of the other agents in that system.
The difference between an intelligent system that serves its occupants and one that surveils, constrains, and extracts from them is not technological. It is a design decision.
Design failure — trauma-exacerbating
Opaque, adversarial systems — driven by data monetization, behavioral channeling, and vendor interests with no mandate for genuine occupant benefit. Operators do not trust them, cannot understand them, cannot opt out without operational cost. Chronic friction. Degraded performance. The technology meant to help becomes the constraint.
Design success — trauma-mitigating
Trustworthy, transparent systems — each with a mandated prime motive of maximizing their human-AI team's success and well-being, minimizing stressors, and supporting the broader mission. Operators understand how they work, trust them, and collaborate with them. Empowered agency. Genuine operational flourishing for all agents in the system.
Design implication
If the tools and environments we design become extensions of the minds and bodies of the people and agents who use them — then designers bear a share of responsibility for those extended minds and bodies. This is not a soft claim about user experience. It is a hard claim about the ethics of the design act. The Context of Use is not neutral. It is either helping agents think and act well, or it is not.
The Context of Use is not a container.
It is an active participant in cognition.
The artifact ecology
An artifact ecology is the constellation of information systems — physical and non-physical, static and interactive, designed and emergent — that surround any cognizing agent and that the agent uses to think, act, and adapt. The concept is grounded in embodied cognition: cognition is not confined to the brain. It is distributed across brain, body, and technological environment. The agent and the artifact ecology are a single extended cognitive system.
Given this, the design of a system's artifact ecology is, in effect, the partial design of the extended minds of everyone who inhabits it. The tools, interfaces, displays, spatial configurations, and adaptive systems that populate an agent's environment do not merely support cognition — they shape it. They prime certain kinds of attention, foreground certain kinds of information, afford certain kinds of action, and constrain others.
The integration of intelligent systems into the Context of Use is not a disruption of this relationship. It is a continuation — at unprecedented speed, scale, and adaptive capacity. What is new are the ethical stakes that accompany that scale. Owners, program managers, and policy-makers who understand this build better systems. Systems that perform cognitively, not just mechanically. Systems that support the teams inside them rather than constraining them.
This is the design challenge that defines Industry 6.0. It requires frameworks, constructs, and measures adequate to its complexity. XPLR Design provides them.
Occupants seed their environments with structures that afford efficient, low-cognitive-load task completion. Good design makes this seeding possible — and makes the resulting structures legible, trustworthy, and adaptable as the system evolves.
Intelligent adaptive systems must be sensitive to the cognitive and physical workflows of the teams operating within them — tuning themselves to the social needs of those teams, not the reverse.
As adaptive systems evolve, the artifact ecology must remain coherent — predictable enough that occupants can model and trust it, adaptive enough to meet the demands of changing operational conditions. Incoherent adaptation produces cognitive overload, not optimization.
Intelligent adaptive systems cannot be fully designed before deployment. The overall design is finished by cultivating the system into existence while in use — with agents, environments, tools, and organizational processes coevolving together across the full system lifecycle.
About XPLR Design, LLC
Our work spans the full human-AI-systems ecosystem — from requirements through operations — bringing the theoretical and empirical foundation of human factors, cognitive science, and systems engineering to bear on the design of intelligent adaptive systems and the teams that operate within them.
For right now, XPLR Design is Joe Manganelli. The organization will evolve beyond that soon.
Key Personnel
Joseph Manganelli
AIA · CHFP · LEED AP BD+C · PhD · Principal Human Factors Researcher
Joe is a practicing registered architect and board-certified human factors professional, as well as a sustainable design specialist. His goal is improving the capacity of the built environment to enhance human cognitive and physical health, well-being, and performance.
Joe's research journey started in architecture school. His thesis focused on the built environment's ability to challenge or affirm people's belief systems, the resultant impact on their psychological well-being, and the implicitly ethical act of environmental design — especially for adaptive environments. This led to an abiding interest in cognitive science, particularly embodied cognition, systems science, and ecological niche construction, as well as industrial architecture.
For more than twenty years, Joe's research and practice integrates constructs and methods from architecture, human factors, systems engineering, cognitive science, evolutionary biology, and sustainability. His human factors practice has focused on industrial, healthcare, and automotive concerns. This background — developed to learn how to better design for humans in task-oriented, mission-critical environments — has led to research developing constructs and methods for human-centered cyber-physical systems, and methods for measuring and analyzing system-of-systems performance of human-AI-robot teams.
Recent & Forthcoming Publications
2026 · Conference Paper · HFES
Manganelli, J. et al. (2026). "Perspectives on Ecological Validity and Real-Time, Multi-sensory Data for Human-AI Systems." Proceedings of the Human Factors and Ergonomics Society Annual Meeting. SAGE Publications.
2026 · Book · CRC Press
Mokhtar, T., & Manganelli, J. (2026). Coexistence and Coevolution of Humans and Intelligent, Adaptive Environments. CRC Press / Taylor & Francis. View on Routledge →
In Development · Standard · ANSI/HFES 900
Human-Artificial Intelligence Systems Test and Evaluation (HAISTE). ANSI/HFES 900 — Draft Standard. Contributing author.
2024 · Invited Presentation · Fluor, Inc.
Manganelli, J. (2024). "Bridging the Gap: Human Factors and Artificial Intelligence in Industrial Automation." Fluor Innovation Builders Series. youtu.be/2vw4XLl9ios
2023 · Conference Paper · IEEE ARSO
Mokhtar, T.H., Manganelli, J., and Hamidalddin, A.A. "A Human-Centered Design Process for Developing Non-Humanoid Social Robotic Work and Exercise Environments." IEEE ARSO 2023, Berlin, pp. 110–115. doi: 10.1109/ARSO56563.2023.10187543. HC Design Process for Developing NHSR →
2018 · Technical Report · NIST
Bhatt, V., Fracsella, A., Brutti, A., Jeong, S., Burns, M., Manganelli, J., et al. (2018). A Consensus Framework for Smart City Architectures. National Institute of Standards and Technology. IES-City Framework →
2018 · Conference Paper · HFES
Burns, M., Manganelli, J., Wollman, D., et al. (2018). "Elaborating the Human Aspect of the NIST Framework for Cyber-Physical Systems." Proceedings of the Human Factors and Ergonomics Society Annual Meeting, Vol. 62(1), pp. 450–454. SAGE Publications. Human Aspect of CPS →
Get in touch
XPLR Design works with government agencies, system owners, and human-AI-systems developers and operators to assess human-AI team performance, develop requirements for intelligent systems, and guide integration across the full project lifecycle.
If you are setting up human-machine or human-AI teams, specifying requirements for AI-enabled systems, or developing policy that governs intelligent systems in the built environment — reach out.
areas of engagement
Requirements engineering for AI-enabled systems
Human-AI team task allocation & levels of automation assessment
Trust calibration and human-AI-systems testing & evaluation
Sociotechnical systems & participatory design
Policy guidance for intelligent systems programs
start the conversation
Send a brief description of your project or initiative. We will respond within one business day.
Send an Email →