Intersubjectivity is conceptualized as a dynamic, embodied process that emerges during growth through multisensory engagement and motor development. The present thesis aims to adapt developmental protocols to ensure inclusive models for infants and children with visual impairment (VI). These adaptations replace gaze-dependent measures with indicators derived from alternative sensory channels, such as touch, proprioception and audition. To investigate how visual experience shapes the reorganization of intersubjective dynamics during early development, the project adopted a multisensory and technology-driven approach articulated across four E-dimensions: Embodied, Embedded, Enculturated and Enabling. Embodied intersubjectivity investigates how visual deprivation is associated with differences in the discrimination of affective touch, revealing preliminary evidence of shifts in caregiver-specific physiological tuning in infants with VI. Embedded intersubjectivity adapts a protocol inspired by the Strange Situation paradigm to children with VI and explores how unexpected multisensory input could modulate kinematic variability and proximity to caregivers, suggesting that flexibility in attention might function as a strategy for environmental control when visual information is limited. Enculturated intersubjectivity develops and validates a multimodal framework for detecting episodes of joint attention through head orientation during naturalistic caregiver and child interactions, integrating clinical video coding, computer vision, and marker-based motion capture. Enabling intersubjectivity presents an innovative sonification system that translates movement into sound in real time, suggesting that auditory feedback can support proprioception, scaffolding motor exploration in the absence of vision. Across these four dimensions, the findings suggest that in infants and children with VI autonomic responses might shift, motor variability could play an adaptive, head orientation provide a potential operational index of joint attention, and auditory cues might scaffold navigation and bodily awareness. The significance of this thesis lies in its strong translational potential. It suggests that interactive research paradigms traditionally rooted in vision can be redesigned around residual sensory abilities, and that technological tools can be used to both detect and support engagement and scaffold navigation. Together, these contributions have the potential to open new avenues for research and clinical practice. A crucial application concerns the transition from qualitative to quantitative assessment of relational indicators, leveraging time-series analysis to capture the dynamic nature of interactive processes. Future developments will aim to extend these metrics to other clinical populations, as relational processes constitute a domain of investigation across neurodevelopmental disorders. Intersubjectivity is far from being a static trait and emerges as a dynamic relational system, offering not only a powerful lens to understand development, but a new way of seeing when vision itself is compromised.