Cognitive and social development emerge from unified neural mechanisms that support predictive processing and active inference. These mechanisms enable individuals to generate expectations about incoming sensory input, regulate internal states, and guide action based on learned internal models. Importantly, such mechanisms are thought to underlie cognitive development across domains, including both sensorimotor coordination and social interaction, within a common predictive framework. These internal models are shaped not in isolation, but through ongoing interactions with the physical, social, and sensory environment. In neurodiverse individuals, variations in these predictive processes and environmental interactions can lead to distinctive developmental trajectories, producing unique sensory profiles, regulation strategies, and expressive behaviors.

A key challenge is the integration and alignment of multimodal signals across neural, physiological, and behavioral domains. While many studies have focused on single techniques- such as EEG, HRV, or gaze - they often fail to capture how these systems interact. As a result, we lack a holistic view of the brain–body–environment loop. New approaches are needed to unify signals across time, methodological approaches, and context, while also accounting for individual variability in how regulatory and expressive patterns are organized. Such integration is essential for studying dynamic co-regulation both within and across individuals in developmentally and contextually meaningful ways.

A second challenge lies in linking these micro-level predictive and regulatory mechanisms to macro-level developmental outcomes. Even when multimodal signals are available, we often lack models that can explain how these signals give rise to complex behaviors such as joint attention, co-regulation, and participation. Capturing these developmental cascades across timescales and contexts is essential for understanding both typical and neurodiverse trajectories. Computational modeling can serve as a methodological bridge to simulate, test, and interpret such cross-level dynamics.

A third challenge is translating these scientific insights into embodied, adaptive systems that function in everyday life. Most current systems are developed and evaluated in controlled settings, disconnected from the lived experiences of neurodivergent individuals. Translational work is needed to bridge the gap between mechanism and implementation by developing robots, wearables, and interfaces that are personalized, context-aware, and co-designed with users, and that operate meaningfully in varied settings.

This Research Topic centres on efficiently translating insights from neurodiverse developmental and social circuits into embodied systems that support real-world communication and participation. The aim is to address key aspects of translating neurodiversity from predictive and developmental processes into embodied machines to deepen understanding of neurodiverse functioning in naturalistic settings.

We welcome studies that (i) address challenges of aligning and integrating multimodal signals across neural, physiological, and behavioral domains—while accounting for individual variability; (ii) link micro-level predictive and regulatory mechanisms to macro-level behaviors such as co-regulation, emotional expression, and participation, using computational modeling as a methodological bridge; and (iii) develop and evaluate embodied, adaptive systems—such as robots, wearables, or AR interfaces—that are personalized, context-aware, co-designed with users, and responsive to everyday real-world contexts.

To gather further insights in translating neurodiversity to embodied machines, we welcome articles focusing on:

Signal integration and individual variability
• Studies investigating predictive or regulatory mechanisms through unimodal signals such as EEG, HRV, respiration, gaze, or prosody.
• Multimodal integration of neural, physiological, and behavioral signals, including methods for temporal alignment, contextual synchronization and computational modeling of the wide gamut of information present in naturalistic stimulations.
• Analyses of individual variability in signal dynamics across developmental stages, interaction contexts, or neurodivergent profiles.
• Research characterizing dynamic co-regulation and signal coupling within the brain–body–environment loop.
Linking micro-level signals to developmental outcomes
• Studies that examine how predictive or regulatory signals give rise to behaviors such as joint attention, emotion regulation, and social participation.
• Computational modeling (e.g., state-space models, RNNs, transformers) used to simulate cascades, predict interaction outcomes, or test developmental hypotheses.
• Research that connects within-individual dynamics to interaction-level or longitudinal developmental patterns.
• Theoretical and empirical studies modeling how predictive processing mechanisms unfold across timescales and levels of behavior.
Application and translational
• Evaluation of embodied systems (robots, wearables, AR) that support neurodiverse users through co-regulation, feedback, or adaptive interaction.
• Development of personalized and context-aware control systems grounded in predictive processing or physiological signals.
• Studies applying participatory design and accessibility principles to AI-supported systems for education, therapy, or everyday life.
• Translational deployments that assess usability, co-designed adaptation, or ethical implementation in real-world environments.

Are you interested in contributing a manuscript to this Research Topic?

Translating neurodiversity: From predictive mechanisms to embodied systems