SensoryNeural Project

Competition Recognition

SensoryNeural earned second place at the Elevate Great AI Competition 2025. The organising team highlighted our concept in their official LinkedIn recap, thanking the judges and celebrating the finalists who helped make the event a success.

Finalists applauded in the LinkedIn announcement:

Chujie Sun
Yunlin Liang
Wenxuan Xu
Hanrui Wu
Taofeek Abimbolu
Godwin Mwisomba
Toheeb Ajala Husain
Imaan Soliman
Ryan Gichuru
Goodness Somtochukwu Odiaka

The Challenge

Neurodivergent children, particularly those with autism spectrum disorder (ASD) and sensory processing disorders (SPD), experience significant distress due to sensory dysregulation. This distress presents as sensory hyper-responsivity, hypo-responsivity, or sensory-seeking behaviors, leading to challenges in:

Social interaction and communication
Emotional regulation and anxiety management
Academic performance and learning
Daily functioning and independence

Traditional interventions like occupational therapy or static sensory tools (e.g., noise-cancelling headphones) fail to address these sensory triggers in dynamic, real-world environments, creating a critical gap in support for these children and their families.

Our Vision

SensoryNeural represents our vision for an adaptive environment that supports neurodivergent children's sensory needs. By combining artificial intelligence with everyday surroundings, we aim to create spaces that adjust to the child rather than requiring the child to adjust to rigid environments.

Our approach shifts from reactive interventions to proactive support, leveraging emerging technologies to predict and address sensory challenges before they lead to distress. The goal is to empower neurodivergent children with greater autonomy, inclusion, and overall well-being.

Our ultimate vision is to shape an inclusive future where neurodivergent children have the same opportunities to thrive as their neurotypical peers.

Conceptual Features

Predictive Sensory Support

Anticipating potential sensory triggers before they cause distress, allowing for timely interventions.

Environmental Adaptations

Creating sensory-friendly spaces that can adjust in real-time to meet individual needs.

Personalized Approach

Recognizing that each child has unique sensory preferences and needs that evolve over time.

Multi-Environment Support

Extending sensory accommodations beyond the home to educational and community settings.

Data-Informed Insights

Providing parents and professionals with meaningful information to better understand and support the child.

Promoting Independence

Helping children develop self-regulation skills while providing necessary environmental support.

From wearable data to an actionable signal

For the finals prototype, we used synthetic calibration data shaped around paediatric heart-rate variability. The aim was not to diagnose a child or infer a medical condition: it was to turn a changing input stream into a transparent, prototype-level indication that could prompt a supportive option in the dashboard.

1. Input stream

Synthetic wearable-style heart-rate and variability readings arrive over the prototype's real-time connection.

2. Calibration

Age-sensitive baseline ranges and recent changes are used to contextualise the incoming values.

3. AI-guided signal

Heuristics and model-guided logic combine those patterns into an interpretable stress-support signal.

4. Support surface

The dashboard visualises the signal and can surface a timely, human-reviewed sensory-support suggestion.

Prototype boundary: the demo used synthetic data and was built for concept exploration. It is not a clinical monitoring system, medical device, diagnostic model, or substitute for professional judgement.

Where AI fits

AI was used as a decision-support layer rather than an autonomous authority. It helped connect a live signal pattern to a small, understandable set of possible sensory supports, while the interface kept the underlying status visible to the people using it.

Pattern awareness: detect changes against a calibrated baseline instead of reacting to a single raw reading.
Contextual suggestions: map a prototype-level signal to practical, low-risk support options in the dashboard.
Human-centred presentation: make the reasoning legible to caregivers and educators instead of presenting an opaque score as a verdict.
Responsible prototyping: keep privacy, safeguarding, and the limits of synthetic data visible throughout the concept.

Development Approach

Our project is following a thoughtful, staged development process with an emphasis on privacy, security, and ethical considerations at every step. We're working closely with clinical experts to ensure our approach is grounded in evidence-based practice.

Research & Understanding

Exploring the needs of neurodivergent children through literature review and expert consultation

Concept Development

Creating the framework for our innovative approach to sensory support

Testing & Refinement

Iterative improvement with feedback from experts and potential users

Implementation

Bringing the solution to those who can benefit most

My Role

Ryan Gichuru - AI Specialist

As the AI specialist on this project, I led most of the technical delivery: Python backend prototypes, a stress‑detection loop with synthetic calibration datasets, and a WebSocket‑driven dashboard for real‑time visualisation and interventions. I also scaffolded the TypeScript/React site (Vite + Tailwind), organised run books for reproducibility, and set up the repository structure for production builds. My role involves:

Designing and validating signal processing and model‑guided heuristics for stress detection
Building dashboards and data pipelines to surface actionable insights for caregivers
Collaborating with clinical experts to align safety, privacy and ethical guardrails
Translating complex technical concepts into accessible features for everyday use

This project aligns perfectly with my passion for developing AI solutions that have meaningful human impact, especially for those who are often underserved by traditional technologies.

The Team

Our interdisciplinary team brings together expertise in child psychiatry, artificial intelligence, software engineering, and hardware integration. This diverse skillset allows us to approach the challenge of sensory support from multiple perspectives, ensuring a comprehensive solution.

The team includes professionals with backgrounds in:

Child and adolescent mental health
Artificial intelligence and machine learning
Software development and user experience design
Data analytics and privacy engineering
Hardware integration and IoT systems

Recognition

SensoryNeural was selected as a finalist in the first rounds of the Elevate Great AI Competition 2025, which focused on enhancing early childhood development through artificial intelligence. Our team's innovative approach to supporting neurodivergent children earned recognition from judges.

The competition challenged participants to envision how AI can positively influence key aspects of early childhood development, which allowed us to showcase our vision for creating adaptive environments for children with sensory processing challenges.

Impact Potential

We believe SensoryNeural has the potential to transform daily life for:

Children with sensory processing challenges: Increasing comfort, reducing anxiety, and improving engagement in educational and social activities
Families: Reducing stress and providing tools to better support their children's unique needs
Educators: Enabling more inclusive classroom environments that accommodate diverse sensory profiles
Healthcare providers: Offering new insights and approaches to supporting children with sensory regulation challenges

By addressing sensory needs proactively, we aim to help remove barriers to participation and learning, ultimately contributing to better outcomes for neurodivergent children.

SensoryNeural: Empowering Neurodivergent Children through AI-Driven Sensory Support