Deployable polyhedral structures offer significant potential for emergency shelter systems in post-disaster contexts, where rapid deployment, compact transportability, and scalable spatial coverage are critical for protective infrastructure. Disaster risk is increasing globally, with hazard events becoming more intense and frequent. However, existing design approaches for deployable emergency shelters lack systematic computational frameworks for generating and controlling tetrahedral-octahedral configurations with telescopic actuation, thereby limiting the exploration of deployment strategies for emergency response applications. This research develops a parametric framework for the automated design of deployable tetrahedral-octahedral systems that transform from compact stowed configurations to expanded operational states through the adjustment of telescopic elements. The Rhino/Grasshopper computational framework automates both geometric generation and telescopic rod length control throughout deployment sequences, validating three deployment strategies through differential telescopic control where octahedral members deploy to shorter lengths than tetrahedral members for arch configurations. This research establishes a systematic foundation for exploring deployment behaviors and scalability in emergency shelters and other rapidly deployed structural applications.

The escalating demand for high-performance buildings necessitates the rigorous integration of interdependent systems, including structure, envelope, and environmental services. However, the Architecture, Engineering, and Construction (AEC) industry continues to rely on fragmented, sequential workflows that fail to manage complex interdisciplinary trade-offs. This disconnect creates a critical barrier to achieving aggressive sustainability goals, as ad-hoc communication cannot substitute for mathematical system coordination. This paper proposes a formal systems engineering framework based on Analytical Target Cascading (ATC) to bridge the gap between architectural intent and engineering optimization. ATC is a hierarchical, decentralized optimization method that decomposes complex systems into manageable subsystems (System → Subsystem → Component). Unlike traditional "All-at-Once" (AAO) centralized optimization, which seeks a global optimum by solving all variables simultaneously, ATC allows subsystems to optimize locally while negotiating targets and responses to ensure system-level consistency. We validate this framework through a computational pilot study involving the structural optimization of a steel frame. The study benchmarks a sequential ATC workflow against a centralized AAO Genetic Algo-rithm. The results demonstrate a fundamental trade-off: while the AAO method theoretically identifies the global optimum, it is computationally prohibitive and prone to constraint violations in high-dimensional spaces. In contrast, the ATC framework converged on a fully feasible solution with a 98% reduction in computational cost. Although the sequential implementation of ATC introduced path dependencies that led to a local optimum (a slightly heavier structure), the method’s superior efficiency and robustness confirm its potential for scaling. The paper concludes that ATC provides a viable alternative to the industry’s disjointed design processes, paving the way for future parallelized coordination strategies that integrate structure, energy, and life-cycle costs.

This paper examines how artificial intelligence can support the detection and analysis of urban informality, spaces that remain partially or entirely invisible in official urban datasets. Focusing on three distinct sites, Dharavi in Mumbai, the self-built peripheries of Lima, and Skid Row in Los Angeles, the study investigates how AI can help uncover spatial patterns that challenge normative assumptions about how cities are planned, monitored, and governed.

Using image segmentation applied to high-resolution satellite imagery, the research identifies visual indicators of informality such as irregular lot configurations, ad hoc construction patterns, and infrastructural gaps. In Dharavi, high-density layering complicates the segmentation process, requiring careful contextual calibration. In Lima, the informal expansion along urban edges illustrates how topography and land tenure shape visual markers of exclusion. Skid Row presents a different case altogether: here, the issue is not physical informality but the visibility of social precarity within a hyper-surveilled urban core. These case studies reveal both the strengths and limitations of AI when applied across different types of informal conditions.

By comparing how informality manifests across three urban regions, the study highlights the uneven geographies of invisibility and the risks of flattening local complexity through global tools. At the same time, it shows how AI, when used critically, can support more nuanced understandings of spatial inequality, illuminating gaps in data, policy, and recognition. This research contributes to ongoing conversations on the ethics of machine vision, and how localized insight can inform the global development of urban AI tools.