​Reading acquisition is a fundamental goal of the education system. However, in England, 1 in 4 eleven-year-olds leave school unable to read at the minimum expected level (DfE, 2022), and The Literacy Trust reports a decline in children reading for pleasure. Reading difficulties cause wide-reaching problems, with illiteracy costing the UK approximately £36 million annually in lost productivity, lower earnings, and increased social welfare and healthcare expenses (World Literacy Foundation, 2018).

Phonemic awareness (PA) at school entry - the ability to isolate, segment, and blend individual speech sounds in words - is identified as the leading predictor of reading success, even above intelligence (NRP, 2000). The International Dyslexia Association emphasises that children with weak PA, the core deficit among dyslexic learners, struggle with letter-sound relationships, causing reading difficulties. The Early Years Foundation Stage (EYFS) pre-school framework sets learning standards, specifying the need for explicit PA instruction before school entry. However, it can be difficult to screen very young children for PA, especially to ascertain if pre-verbal children can hear speech sounds in the right order and blend them together. Educators may wait until children can articulate the speech sounds or are learning phonics (the relationship between sounds and letters) causing a delay in effective diagnosis, data collection, tracking, and starting points for instruction.

According to the DfE (2018), 1 in 4 children start school without PA, highlighting the discrepancy between expectations and outcomes. Emma Hartnell-Baker, a neurodivergent woman with an extensive background as an 'Outstanding' nursery owner, OFSTED Inspector, and SpLD specialist teacher trainer, is a doctoral student researching PA and offers 1:1 screening and intervention. Her innovative approach utilises 'Phonemies' - fun International Phonetic Alphabet (IPA) aligned speech sound characters with no link to phonics - used in the Innovate UK-funded project MySpeekie, a one-screen communication device she designed for non-speaking autistic 2-4-year-olds.

With engaging, inclusive, and developmentally appropriate tech-driven screening, we identify more at-risk 3-year-olds and can analyse learning behaviour with greater depth by utilising AI. Screening again at age 4 gives us a clearer picture of 'what works.' Failing to support children to reach standard levels of literacy could result in lifetime costs to society of £830 million for each cohort of five-year-olds (Pro Bono Economics, 2021). This technology, to be commercialised through a licensing model, will help send all children to school with good PA and ensure that parents of identified at-risk children are also supported.

This project addresses an evidenced societal challenge: the early identification of phonemic awareness deficits in good time to address this and prevent reading difficulties. Remedial work in school is far costlier. We aim to become gamechangers in this field, with the potential to lead in neuro-affirming innovation. The funding and support from this award are pivotal for measuring success, and scaling up.

Target customers include UK early years providers, speech therapists, and educational psychologists, who will be able to purchase via a licensing model at an anticipated £75 per child per screening. Insufficient early intervention will generate economic costs of around £7,800 per child (PBE 2024). There are around 3,500 3-year-olds and 3,500 4-year-olds to screen in Dorset alone.

Research shows that screening 3-year-olds is often challenging, time-consuming, and collecting accurate data is difficult (Greninger, 2021; Frontiers in Pediatrics, 2021). Children love our 'Speech Sound Monsters', currently used (human-led) to effectively screen non-speaking or pre-verbal young children for the ability to hear, order, and blend phonemes (sounds) in words using our own phonemic awareness activities. Our innovation, a neuro-inclusive tech-driven PA screening tool, will replicate these 1:1 screening activities on the touchscreen, using AI to better analyse learning behaviours, providing a consistent, scalable solution that gathers extensive data, aiding early intervention to reduce reading difficulties. Assessment is vital, with all children included. This innovation is designed specifically for non-verbal and neurodivergent 3-year-old children, effective primarily because of their interest in Phonemies, and that no prior knowledge of letters are required.

Phonemies are directly linked to 'phonemes' (not phonics) as an alternative to the International Phonetic Alphabet (IPA) symbols and Received Pronunciation (RP) used by speech therapists. They feature in the MySpeekie app, developed with Innovate UK grant funding, as the world's first 'one-screen' AAC device for nonverbal autistic children. We are committed to EDI and are building a unique advisory group focused on solutions that embrace differences and child-centric learning.

Post-award, we will establish a Learning Differences Research and Training Centre to develop and disseminate innovative educational and 'visible speech' therapy-related tools for children from birth, advocating for neuro-affirming practices and preventative approaches. The 'Phonemies' uniquely bridge spoken and written English: already helping toddlers learn to read 'naturally' while they learn to speak. Additionally, we aim to create training courses and jobs for 'visible speech' specialists. We have plans to create an AI reading teacher for children without access to electricity, the internet, or trained teachers, partnering with organisations like the World Literacy Foundation to immunise children against illiteracy globally.