From reacting to predicting.

“When my son’s heart rate jumps about 10 beats a minute within 30 to 45 seconds, a meltdown is very likely coming. Meltdowns can be violent and frightening for everyone. If the adult with him gets an alert in that window, they can redirect him before it starts. That is the moment we are building for.”

The science says the signal is there

• 1 minearly warning of aggression toward others from a wrist biosensor, in a pilot with 20 autistic youths (AUC 0.71 overall, 0.84 for person-specific models) (Goodwin et al., 2019)
• 3 minadvance prediction of aggressive behavior in 70 autistic youths across four inpatient hospitals (AUROC 0.80) (Imbiriba et al., 2023)
• 1 in 31U.S. 8-year-olds has autism (Shaw et al., 2025)
• 40–80%of autistic children are estimated to have sleep problems, especially insomnia (Cortesi et al., 2010)

Published research, not FPK results. AUC and AUROC are measures of how well a model separates events from non-events, where 0.5 is chance and 1.0 is perfect; they are not accuracy percentages. Both prediction studies used research sensors with youths in specialist inpatient settings; results at home and school still have to be shown.

How it will work

1. Wear

   Any compatible watch, band or clip, chosen for the child’s sensory needs.

2. Learn

   FPK-X learns this child’s normal heart rate, movement and sleep.

3. Detect

   A spike while still, not while playing, flags rising stress.

4. Alert

   The parent, aide or therapist on duty gets a tiered alert.

5. Log

   They record what happened and what helped, in seconds.

6. Improve

   Every outcome makes the next prediction for this child sharper.

Why FPK-X, not a device company

• What a wearable sees

  Heart rate, movement, sleep, skin temperature. A signal, with no idea what it means for this child.

• What only FPK-X knows

  The incidents, triggers, medications, meals, school day and “what helped”, logged by the whole team.

• Labeled outcomes

  Body signals matched to real events. That is the data predictive models need, and it grows with every family.

Beyond meltdowns: what the same data can unlock

• Sleep · night waking · restlessness

  Sleep to next-day forecast

  A rough night becomes a morning heads-up to school: “5 hours, woke 3 times, plan a lighter morning.”

  Builds on the FPK-X sleep log

• Sleep · resting heart rate · activity

  Medication response

  Before-and-after views when a medication or dose changes, including afternoon wear-off. Reports for the prescriber, never dosing advice.

  Builds on the medication hub

• Location · heart rate and motion surge

  Elopement safety

  Geofence alerts paired with a stress surge. In a U.S. survey of parents of 1,218 autistic children, 49% said their child had tried to bolt after age 4. Of the children who went missing long enough to cause concern, 65% were in danger of traffic injury (Anderson et al., 2012).

  Builds on incident reports

• Heart-rate recovery time

  Calming-strategy scorecard

  Shows which strategies bring this child back to baseline fastest, so the team uses what actually works.

  Builds on “what helped” logs and behavior plans

• All signals, over time

  Objective evidence

  Body-signal data alongside staff logs for IEPs, behavior assessments and health-plan medical-necessity reviews.

  Builds on FPK-X reports

• Exploring

  Seizure alerts

  About 1 in 10 autistic people has epilepsy (Liu et al., 2022). We plan to connect to FDA-cleared seizure-alert devices, not to claim detection ourselves.

• Exploring

  Illness and pain warning

  In a study of 32 adults with COVID-19, smartwatch data showed changes in 81%, and in 22 of 25 cases with symptom data the changes came before or at symptom onset, a median of 4 days early (Mishra et al., 2020). It matters most for children who can’t say they feel unwell.

Built responsibly

• Consent and assent

  Parent consent and the child’s assent come first. The family decides who receives alerts.

• Decision support, not diagnosis

  Alerts help the adult who is there decide what to do. They do not diagnose, and they never replace medical advice.

• Regulatory review before launch

  We will complete regulatory review before release, and we won’t claim to detect medical events such as seizures ourselves.

• Privacy from the start

  Body-signal data sits in the same permissioned FPK-X record, with the same role-based access.

Path to launch

1. Today

   FPK-X is live with schools and families.

2. Build

   Device-agnostic integration and per-child baselines.

3. Q2 2027

   Release to early-access families.

4. Next

   Schools and new features, explored with partners.

Plan, subject to change.

Start with autism. Then every setting.

Autism is where coordination is hardest and the need is clearest, so it is where we prove the model. The same engine of personal baselines, labeled outcomes and alerts to the right person carries into the settings that follow: disability and independent living, mental health, elder care, chronic conditions, post-surgery recovery, athlete performance and high-performance training. See every use case.

References

1. Anderson, C., Law, J. K., Daniels, A., Rice, C., Mandell, D. S., Hagopian, L., & Law, P. A. (2012). Occurrence and family impact of elopement in children with autism spectrum disorders. Pediatrics, 130(5), 870–877. doi:10.1542/peds.2012-0762
2. Cortesi, F., Giannotti, F., Ivanenko, A., & Johnson, K. (2010). Sleep in children with autistic spectrum disorder. Sleep Medicine, 11(7), 659–664. doi:10.1016/j.sleep.2010.01.010
3. Goodwin, M. S., Mazefsky, C. A., Ioannidis, S., Erdogmus, D., & Siegel, M. (2019). Predicting aggression to others in youth with autism using a wearable biosensor. Autism Research, 12(8), 1286–1296. doi:10.1002/aur.2151
4. Imbiriba, T., Demirkaya, A., Singh, A., Erdogmus, D., & Goodwin, M. S. (2023). Wearable biosensing to predict imminent aggressive behavior in psychiatric inpatient youths with autism. JAMA Network Open, 6(12), e2348898. doi:10.1001/jamanetworkopen.2023.48898
5. Liu, X., Sun, X., Sun, C., Zou, M., Chen, Y., Huang, J., Wu, L., & Chen, W.-X. (2022). Prevalence of epilepsy in autism spectrum disorders: A systematic review and meta-analysis. Autism, 26(1), 33–50. doi:10.1177/13623613211045029
6. Mishra, T., Wang, M., Metwally, A. A., et al. (2020). Pre-symptomatic detection of COVID-19 from smartwatch data. Nature Biomedical Engineering, 4(12), 1208–1220. doi:10.1038/s41551-020-00640-6
7. Shaw, K. A., Williams, S., Patrick, M. E., et al. (2025). Prevalence and early identification of autism spectrum disorder among children aged 4 and 8 years, ADDM Network, 16 sites, United States, 2022. MMWR Surveillance Summaries, 74(2), 1–22. doi:10.15585/mmwr.ss7402a1