CuddlSense
A soft plush companion for children with sensory processing differences that recognises how they squeeze, pat, or hit it and responds with matched calming or alerting feedback for parents and therapists.
Parents of children aged 3–12 with autism, ADHD, or sensory processing needs, working with an OT
- Real-time touch-style recognition (pat, squeeze, hit, cuddle) via embedded sensors at 20 Hz
- Parent and therapist dashboard surfacing weekly patterns — 'squeezing spiked on Monday mornings'
- Customisable response (vibration, warmth, light, gentle audio) matched to the detected affect
- End-of-day summary that flags any high-force negative touch sessions for safeguarding follow-up
The paper validates that a 13k-parameter 1D CNN can run at 20 Hz on a microcontroller and detect subtle social touches better than heuristics, making an actually-affective plush companion technically feasible for a small hardware team for the first time.
Established market for autism/ADHD sensory products (weighted blankets, lap pads, plush) plus emerging AI plush category (Amazon, Ollobot, Ropet, PingoPongo) targeting ages 3-12, but this remains a niche within niche.Autism and Plushies: Sensory and Emotional Support ↗Amazon.com: Ai Stuffed Animals ↗
PingoPongo (SEPA Babies) is a near-direct competitor — smart plush companion supporting emotion regulation for neurodiverse children with on-device AI; Chinese AI plush toys (Ollobot, Ropet) also encroaching; the specific affective-touch-via-1D-CNN angle is novel but defensibility is unclear.SEPA Babies — Meet PingoPongo, the Smart Plush Companion ↗Best AI Plush Toys 2026: We Tested 7 Chinese AI Pals ↗
Sensory/therapeutic products command premiums ($50–$200 weighted items, $150–$300 AI plush); parent demographic for autistic kids spends willingly but is price-sensitive on non-reimbursed items and COPPA compliance adds cost; subscription/app models plausible.Best Weighted Blankets for Children with Autism & ADHD ↗Best AI Stuffed Animal | Soft AI Pets That Hug Back & Learn ↗
Rising autism/ADHD diagnoses and growing OT-driven awareness of sensory needs support a long-running category; arXiv validation paper and 1D-CNN-on-microcontroller trend signal durable tech tailwind rather than hype.Design and Validation of a Lightweight 1D CNN for Affective Touch Classification in Soft Plush Companions ↗
Technical feasibility validated by the arXiv paper (13k params, 20 Hz on MCU), but it's still a child-facing hardware product with COPPA, FCC, and textile-safety requirements; small hardware team realistically needs 12–18 months and meaningful capital.Design and Validation of a Lightweight 1D CNN for Affective Touch Classification in Soft Plush Companions ↗