Children on the autism spectrum frequently require repeated, structured opportunities to practise the communication skills introduced during speech pathology sessions. Access to therapy is constrained by cost, NDIS funding limits and session frequency, creating a gap between appointments during which skill reinforcement is limited. Socially assistive robots offer a promising mechanism for bridging this gap, yet existing platforms rely heavily on pre-scripted interactions, lack personalisation and require close operator involvement, limiting real-world scalability.
This project addresses those limitations by combining a physical robot platform (Reachy Mini) with a generative AI backend grounded through retrieval-augmented generation (RAG). Rather than operating autonomously, the system is designed as a clinician-centred tool: therapists define lesson goals, approve interaction content and review session outputs via a purpose-built dashboard. Parent and guardian input is incorporated to contextualise robot prompts with recent real-world events. Session audio from clinical appointments is transcribed and analysed to generate lesson plans that guide future Reachy sessions, creating a feedback loop between formal therapy and between-session practice.
Aim
The central aim of this project is to design, develop and evaluate a clinician-guided generative AI-driven socially assistive robot system that measurably supports communication skill development in autistic children between scheduled therapy sessions.
Objectives
- Design and validate a structured interaction framework for the Reachy Mini capable of delivering clinician-approved, goal-directed communication tasks.
- Develop a RAG pipeline that personalises robot interactions using client profiles, session transcripts, clinician notes and parent/guardian input.
- Implement and evaluate clinician-facing and parent/guardian-facing dashboards that maintain clinical oversight and data integrity.
- Establish and apply a formal evaluation protocol measuring communication outcomes (mean length of utterance, vocabulary growth, verb complexity, narrative macrostructure) and system usability.
- Conduct real-world feasibility testing with clinician and family participation to assess therapeutic alignment and practical adoption.
Significance
Autism affects approximately 1.1% of Australians, rising to 4.3% among school-aged children, with the majority requiring additional support beyond current provision. This project addresses a clinically significant gap in between-session practice, leveraging advances in generative AI and social robotics to make structured, personalised communication support more accessible. The system’s clinician-centred design distinguishes it from prior autonomous robot platforms and directly responds to evidence that human oversight maximises therapeutic outcomes in robot-mediated intervention. Findings will contribute to the fields of socially assistive robotics, AI in clinical support tools and autism intervention design.
Ideal Candidate
The ideal candidate will hold an honours or master’s degree in mechatronic engineering, computer science, human-computer interaction or a related field. Strong Python programming skills are essential, with experience in NLP, large language models or machine learning highly desirable. Familiarity with human-robot interaction, assistive technology or clinical research contexts will be advantageous. The candidate should demonstrate the ability to work across technical and clinical domains, communicate effectively with non-technical stakeholders and manage a multidisciplinary research project. Experience with database systems, speech processing or evaluation methodology would be beneficial. Additionally, the applicants should meet the eligibility criteria for entry into a PhD program at Curtin University.
This project is open to Domestic applicants only.
Scholarship
If you are identified as the preferred candidate for this project, you may be considered for an RTP scholarship.
Enquires and How to Apply
For enquires about this opportunity contact Associate Professor Susannah Soon at Susannah.soon@curtin.edu.au
To formally apply submit an Expression of Interest to Associate Professor Susannah Soon during the Central Scholarship round (July 1st – July 31st 2026)