Brooks Band · Family courts, CPS & foster care
Context, not conclusions.
The flashlight for the rooms no one sees.
Brooks Band is being explored as a way to provide authorized professionals with selected baseline-derived measurements and recovery context alongside observations, records and the full evidentiary picture. The prototype is not court validated and does not determine safety, credibility, parenting quality or custody. The data informs; qualified people decide.
Court and child-welfare workflows are a development and validation pathway. No court pilots have run yet; formal use requires technical validation and jurisdiction-specific legal review.
The system is asked to decide with very little to see.
4.7 million
referrals to child protective services in federal fiscal year 2024.
2.9 million
children received a CPS investigation or alternative response in 2024 (2,900,234).
About 8 in 10
of those children were not found to be victims (532,228 victims were identified).
Calculated from HHS national figures.
1,773
children died from abuse or neglect in 2024, as reported by states.
343,077
children were in foster care on September 30, 2023, with 175,283 entries that year.
4 to 8 children a day
are estimated to die from abuse and neglect.
Commission to Eliminate Child Abuse and Neglect Fatalities, Within Our Reach (2016)
Same gap in every room: a 30-minute visit is seen. The other 167½ hours of the week are not.
Who it serves
One record. Each role sees its own view.
Judges & magistrates
Their question
What does the neutral record show across both homes and over time?
What they can see
A proposed view of selected timestamped trends, only with appropriate consent or lawful authority and validation.
What they never see
Audio, words, locations of conversations, scores or recommendations.
CPS & child welfare caseworkers
Their question
How is this child's baseline changing between visits?
What they can see
A proposed view of selected trends between home visits, alongside direct observations and other records.
What they never see
Recordings of the family, or any conclusion about an allegation.
Guardians ad litem & CASA volunteers
Their question
What selected patterns may deserve follow-up?
What they can see
A developing context view, not proof of what happened or a substitute for investigation.
What they never see
Private conversations, or anything the court has not authorized.
Foster parents & kinship caregivers
Their question
How is this child settling in?
What they can see
A proposed view of the child's own baseline and personally appropriate support goals.
What they never see
Records from other placements unless the agency authorizes it.
Parents (either household)
Their question
What routines or support may help my child?
What they can see
Selected derived context authorized for their household, without parenting scores or blame.
What they never see
The other household's data unless ordered.
Therapists & evaluators
Their question
Is what I see in the office consistent with the rest of the week?
What they can see
Developing baseline summaries for professional interpretation alongside their own observations.
What they never see
Anything they write for them — professionals write their own documentation.
Where it helps
The weeks between the visits.
Before and after a custody exchange
A future authorized view could compare selected measurements around transitions. A trend would not establish why a change occurred or determine custody.
A contested allegation
Selected derived records could add context alongside the full evidentiary record. They cannot establish whether an allegation is true or false; routine records are not conversation transcripts.
A foster placement
A future authorized view could explore change-from-self and recovery patterns across routines, without turning the device into a placement score.
When explaining is difficult
Individualized baseline context and support preferences may offer a starting point for understanding. They do not tell others what a child thinks or feels.
Context from the room
Microphone-derived sound measurements can add environmental context; routine records use derived measurements rather than conversation transcripts.
What it is
- A pre-release development and validation pathway for professional review
- Each child compared only to their own baseline
- Selected timestamped trends with audit controls still being developed
- Designed for appropriate consent or lawful authority and role-limited access
- Built on Samsung Galaxy hardware
What it is not
- A truthfulness tool — it never judges credibility or what was said
- A decision-maker — no scores, verdicts or recommendations
- A medical device — general wellness only
- A conversation transcript service — routine records are derived measurements
- Surveillance of parents or caregivers
- A replacement for an investigation, a home visit or a professional's judgment
How access works
Consent first. Logged always.
1 · Consent or court order
Data is shared only with a legal guardian's consent or under a court order, as the court directs.
2 · Role-based access design
A court-focused deployment would require role-limited access tied to consent or lawful authority.
3 · Audit controls to validate
A court-focused workflow would require records of views, exports and disclosures; these are not confirmed finished features.
4 · Revocable and time-limited
Access can be ended, and data is kept only as long as the retention terms allow.
The research
Why voice and the body: the research
These findings are from other research teams, not Brooks Band results or validation of its court use. They inform research questions about voice and wrist signals. Brooks Band is being developed around a person's own baseline, not medical or legal conclusions.
Voice in young children
Machine learning analysis of a 3-minute storytelling task identified children ages 3–8 with an anxiety or depression diagnosis with 80% accuracy (71 children).
McGinnis et al., IEEE Journal of Biomedical and Health Informatics, 2019.
Wrist signals ahead of an outburst
A wristband measuring heart rate, sweat, skin temperature and movement predicted aggressive episodes about one minute in advance with 84% accuracy (20 children with autism).
Goodwin et al., Autism Research, 2019.
Voice and trauma
Speech-based markers correctly identified veterans with a PTSD diagnosis 89% of the time (52 veterans with PTSD, 77 without).
Marmar et al., Depression and Anxiety, 2019.
Voice pitch and stress
A systematic review and meta-analysis examined the fundamental frequency of the voice (pitch) as a potential stress biomarker.
Veiga et al., Stress and Health, 2025.
Children's speech and heart rate under stress
Researchers found measurable differences in speech behavior and heart rate in relation to task-induced stress in children with and without autism.
Bone et al., Interspeech, 2017.
Legal & regulatory notes
- General wellness product. Brooks Band is pre-release general wellness support technology. Not a diagnosis or a medical device. It is not commercially final or court validated. FDA General Wellness guidance
- The court decides. Whether and how any data is considered in a proceeding is for the court under the applicable rules of evidence. Federal Rule of Evidence 702
- Audio privacy by design. Routine records are designed around derived measurements rather than conversation transcripts. Some transient audio processing may occur for wearer verification, including cloud processing depending on configuration. Recording and consent laws vary; courts and agencies must review requirements with counsel.
- This page is general information, not legal advice.
Last reviewed: September 2026
Sources
- HHS, Child Maltreatment 2024
- HHS, Child Maltreatment 2024 (via The Imprint) — maltreatment reports, victims and fatalities down in 2024
- AFCARS FY2023 (via The Imprint) — number of youth in foster care declines again
- AFCARS data — Children's Bureau adoption and foster care data
- Child Trends, Child Welfare Financing SFY 2022
- Commission to Eliminate Child Abuse and Neglect Fatalities, Within Our Reach (2016)
- McGinnis et al., IEEE Journal of Biomedical and Health Informatics, 2019 — PubMed
- McGinnis et al., 2019 — full text
- McGinnis et al., 2019 — summary
- Goodwin et al., Autism Research, 2019 — journal
- Goodwin et al., 2019 — Northeastern University summary
- Marmar et al., Depression and Anxiety, 2019 — PubMed
- Marmar et al., 2019 — NYU Langone summary
- Veiga et al., Stress and Health, 2025 — full text
- Veiga et al., Stress and Health, 2025 — journal
- Bone et al., Interspeech, 2017 — paper
- Bone et al., Interspeech, 2017 — full text
- FDA, General Wellness: Policy for Low Risk Devices
- Federal Rule of Evidence 702 (Legal Information Institute)
Courts, CPS & foster care
Bring the flashlight to your court or agency.
Please don't include case details, names of children, or confidential information in this form.
Meet with me