Across four decades of AI deployment, the same six human challenges (trust calibration, reliance behavior, cognitive engagement, skill retention, accountability, and transparency) recur, yet fragmentation across research communities obscures this continuity and limits knowledge transfer. Functionally similar phenomena are repeatedly relabeled (a jangle fallacy): what aviation researchers call “automation complacency,” decision scientists call “algorithm appreciation,” and LLM researchers describe as “over-reliance.” This systematic review synthesizes 152 papers spanning aviation, healthcare, manufacturing/supply chain, and cross-domain contexts across three AI technology generations: decision support systems, autonomous systems, and large language model (LLM) agents. We introduce the Collaboration Convergence Framework (CCF), a 6 × 3 matrix with solution-maturity indicators that maps each challenge across generations. The framework shows that Gen 3 designers can transfer decades of evidence from automation and decision support research (particularly reliance calibration, cognitive forcing, and skill maintenance) rather than rediscovering them. Cross-generational synthesis also isolates three Gen 3 phenomena without direct precedent in earlier generations: epistemia (attributing genuine knowledge to LLMs based on surface fluency), attribution ambiguity in co-creation, and motivational withdrawal. We distill twelve transferable design principles and propose ten research directions, prioritizing skill-retention interventions and accountability frameworks. These findings carry direct sustainability implications aligned with Industry 5.0: protecting workforce capability under increasing automation (SDG 8), reducing duplicated research effort through cross-generational knowledge reuse (SDG 9), and supporting responsible deployment by treating collaboration risks as predictable rather than novel (SDG 12). The CCF provides conceptual infrastructure for cumulative learning across AI generations and industries.
Since the introduction of Industry 4.0, digital twin technology has advanced dramatically, and its applications have been extensively discussed. Through digital twins, real-time connected production systems are anticipated to be more efficient, resilient, and sustainable, facilitating communication and connectivity between digital and physical systems. However, challenges remain in environmental performance and in integration of virtual reality (VR) and artificial intelligence (AI). Further exploration of digital twin technologies is needed to validate the real-world impact and benefits. This paper investigates these challenges by implementing a real-time digital twin based on the ISO 23247, Automation Systems and Integration-Digital Twin Framework for Manufacturing, (Digital twin framework for manufacturing), connecting the physical factory and simulation software with VR capabilities. This digital twin system provides cognitive assistance and a user-friendly interface for operators, thereby improving cognitive ergonomics. The connection of the Internet of Things platform allows the digital twin for real-time bidirectional communication, collaboration, monitoring, and assistance. A laboratory-scale drone factory was used as the digital twin application to test and evaluate the ISO 23247 standard as well as its potential benefits. Additionally, AI integration and environmental performance key performance indicators have been considered as the next stages to improve VR-integrated digital twins. This paper further discusses how digital twins contribute to the improvement of the environmental performance of production systems, primarily from the perspective of data inventory management. Supported by a solid theoretical foundation and a laboratory-scale demonstration of the VR-integrated digital twins in a drone framework, the paper bridges the gap between possible technological integrations and the advancement of digital twins framework based on ISO 23247.
Distributed Control Systems (DCS) became ubiquitous in the 1980s and 1990s and remain the backbone of process control in pulp and paper. However, many installations are obsolete and operate without vendor support or reliable spare parts sources, in an environment with rapidly declining institutional knowledge, and burdened by threats relating to OS platforms. Modernization is often deferred as more visible operational issues take precedence. Process control managers are challenged with justifying large-capital DCS projects that seem ambiguously planned and complex to execute that are perceived to provide little or no return on investment.This paper provides (1) a framework for capital justification by defining the value generated by investment, and (2) by identifying factors that contribute to the overall risk and cost relating to the migration of aging control systems. By repositioning a DCS upgrade from an in-kind replacement to an optimization investment, mill teams can improve operational effectiveness. Applying proven methods for the planning and implementation phases of DCS modernizations reduces execution risk and capital cost uncertainty, independent of the selected DCS platform. This paper describes the holistic approach to reducing DCS project risk, and the process to define the project’s measurable value creation.When done well, migrations deliver tangible returns: better control performance and operator efficiency, reduced downtime through improved alarming, diagnostics and troubleshooting tools, and the ability to implement advanced control and optimization.
In the dynamic realm of industrial robotics, prolonged operation amidst vibrations and occasional collisions presents formidable challenges to sensor calibration. Fast and effective methodologies for assessing and remedying calibration discrepancies are imperative to mitigate these issues. In response, we introduce an approach in target-based stereo baseline calibration refinement tailored for real-world industrial environments. Our method offers a means of incrementally assessing accuracy and correcting baseline drift in rectified stereo images. Departing from conventional approaches, our calibration methodology operates under the assumption of an initial stereo camera calibration estimate, focusing exclusively on adjusting minor baseline variations. This streamlined process facilitates efficient enhancement of depth precision, circumventing the need for cumbersome calibration routines. Furthermore, we design a calibration target specifically for this process, combining ArUco markers for precise visual detection with a pseudo-random pattern for robust depth estimation. Through extensive experimentation, we demonstrate the superiority of our method over traditional stereo camera calibration techniques in optimizing depth accuracy. Overall, our work presents a practical solution for maintaining high-fidelity depth estimation in long lived stereo vision systems deployed within industrial settings, underscoring its relevance and applicability in real-world scenarios.
Autonomous systems that build structures from natural-language instructions need reliable spatial reasoning, yet large language models (LLMs) make systematic coordinate errors when generating three-dimensional block placements. We present a neuro-symbolic pipeline based on 2.5-D decomposition: the LLM plans in the two-dimensional horizontal plane while a deterministic executor computes all vertical placement from column occupancy, eliminating an entire class of errors. On the Build What I Mean benchmark (160 rounds), GPT-4o-mini with this pipeline achieves 94.6% mean structural accuracy across 12 independent runs, within 3.0 percentage points of the 97.6% ceiling imposed by architect-agent errors that no builder-side improvement can address. This outperforms both GPT-4o at 90.3% and the best competing system at 76.3%. A controlled ablation confirms that 2.5-D decomposition is the dominant contributor, accounting for 50.7 percentage points of accuracy. The pipeline transfers directly to edge hardware: Nemotron-3 120B running locally on an NVIDIA Jetson Thor AGX matches the cloud result at 94.5% with no prompt modifications. The underlying principle, removing deterministic dimensions from the LLM's output space, applies to any autonomous construction or assembly task where gravity or other physical constraints fix one or more degrees of freedom. A transfer experiment on 500 IGLU collaborative building tasks confirm the effect generalizes beyond the primary benchmark.
