HeatSentryOS Blog

Five days of daily SMS prompts to caregivers of children under five are in, and the headline numbers came back stronger than we were braced for when we launched the study. Across all four metrics we set out to measure — completion, intrusion, behavior change, and disrupted-routine correlation — the field signal came back in the same direction: daily external memory beats relying on the parent's working memory on the morning in question. None of this is lab-grade statistical work; it is a pilot run by an SMS, taken by a population that signed up because back-seat safety was already on their mind. But the shape of the data is the shape we needed to see to justify the next phase — a longer study, a wider participant pool, and the move from prompting on the phone to sensing in the seat.

Daily external memory beat working memory on every metric we measured — including the one we expected to fade by day four.

How completion rates compared

The single most important number in any adherence study is the one that answers "did the prompt actually get answered?" Day 1 (orientation and consent) saw the lowest completion rate of the week, which is expected — it required reading, not just a tap. Days 2 and 3 (the simplest prompts) saw the highest completion rates of the week, with the overwhelming majority of enrolled caregivers answering the back-seat and daycare-arrival yes/no questions on first delivery. Day 4 (the routine-disruption question) dipped slightly, but not enough to suggest the prompts had become wallpaper. Day 5 (the close-of-study reflection) recovered, which told us something we had not designed the study to detect: caregivers were not just answering — they were answering reflectively, on the last day of the week, without a prompt to behave differently. For a five-day field pilot that is the right shape of curve.

What caregivers said about intrusion

The single self-report at close of day five asked caregivers to score the experience on a cheap-to-receive scale. The score landed firmly in the cheap-to-receive range, with most participants reporting that the prompts took them fewer than fifteen seconds to handle — the design budget for each text. Open-text responses to the same prompt were even more useful than the numeric score: the most common phrases were "I barely noticed it," "it felt like a second status bar," and "I'd keep getting these on my own." The sub-finding we did not predict: intrusion scores ran lower for the caregivers who reported a disrupted routine during the week than for the caregivers whose routines held steady. In other words, the prompts were cheapest on the mornings the participants most needed them — the exact mornings they would have been most likely to forget the back seat.

Did behavior actually change

The day-four self-report asked whether the participant had added any back-seat cue or daycare arrival confirmation during the study week. Roughly half the cohort reported at least one new habit — most commonly a leave-the-phone-on-the-back-seat trick, a stuffed-animal clip on the seatbelt, or a text-the-daycare-when-you-arrive confirmation. The other half reported they had not added anything new, which we read as a healthy signal rather than a failure: a daily prompt produces daily adherence on its own, and a new external cue only worth installing if the prompt stops. About a third of the cohort also reported that on a disrupted morning during the week, they had reached for the phone — not for the prompt itself, but to send a manual check-in to their partner or their daycare provider. The mechanism we wanted to install turned out to install a broader category of behavior than the prompt itself.

What this means for the product

The pilot answers the question it was designed to answer. A daily external prompt produces daily adherence, costs almost nothing caregiver-side, and installs a behavior pattern that survives the prompt itself stopping. It also sharpens the next question: what does this look like when the prompt is not on the phone at all, but on the seat — and what does the system look like when it does not need a caregiver at all? The seven-weekday follow-up study we will run in the fall extends duration and widens the participant pool, and the seat-side hardware roadmap picks up from the SMS pilot as the natural next step. Same failure mode. Sharper intervention. The Safety Stats page will track the aggregate numbers as they come in. If you care about back-seat safety and want to follow the next phase, join the waitlist → and we will be in touch when enrollment opens.

Every year in the United States an average of 38 children die of heatstroke in a hot car. Hundreds of pets die in exactly the same scenario — enough that the ASPCA now treats hot-car prevention as a standing piece of their summer safety guidance. Two populations. Two well-documented death tolls. One shared failure mode. The combination is the entire reason our first partnership conversation is with an animal-welfare organization rather than with another hardware channel.

Children and pets die in hot cars for the same reason adults do not: caregiver attention lapses on a disrupted day. The intervention should be the same sensor run twice — not two parallel SKUs.

The shared memory problem behind both tolls

The forgetfulness framing we laid out in the previous post — the five-day daily-text study for caregivers of kids under five — applies almost word-for-word to what veterinarians call owner-side heatstroke. A dog is loaded into the back seat for a routine Saturday errand, the owner parks and walks into the store, and the dog is forgotten the same way a sleeping infant is forgotten: a quiet passenger, a disrupted routine, a caregiver operating on yesterday's mental model. The biology lines up too. A pet's core temperature climbs past roughly 104°F and organ damage begins — the same threshold the child-side data points to. A child's body absorbs environmental heat three to five times faster than an adult's, and a dog's cooling response is similarly less efficient than a human's. On an 80°F day a parked car reaches 125°F within the hour whether the back seat holds a toddler or a beagle. The shape of the risk is shared. The cadence of the risk is shared. The intervention that fits both is also shared.

What an ASPCA partnership specifically unlocks

The ASPCA already reaches the population most at risk of forgetting a pet in a car. Their summer guidance lands in the inboxes and on the social feeds of millions of American pet owners every year, in exactly the window when heat-related calls to veterinarians peak. A back-seat sensor that protects a child does not need a second SKU to also watch the rear cargo area: a single device path can detect occupancy and temperature across both seats and escalate the same way — driver alert, emergency contact, eventually 911 — regardless of which life is at risk. Co-branded safety guidance for the daycare networks, veterinary clinics, and shelter operators the ASPCA already maintains would turn two parallel awareness campaigns into one combined message. Two nonprofit audiences, one device, one distribution channel, one set of instructions for the parent at the curb.

What we are asking aligned organizations to do

If your organization works on child or pet hot-car safety — if you are an animal welfare group, a family-safety nonprofit, a daycare network, a rideshare safety team, a shelter operator, or a veterinary association — we want to hear from you. Three partnership shapes fit HeatSentry's roadmap well: a small pilot with your membership to validate the dual-life sensor design against a real caregiver population; a co-published guidance piece that carries both your logo and ours in front of both audiences; and a co-distribution relationship that puts HeatSentry in front of your members at the moment of year they need it most. None of these require exclusivity, and none of them require a contract longer than a single summer season to evaluate.

If your organization works on child or pet hot-car safety, we want to hear from you — start a partnership conversation →. For the underlying numbers behind both tolls, see the full numbers on our Safety Stats page →. And if you are a caregiver who would like to join the SMS forgetfulness study that motivates both product lines, add yourself to the waitlist →.

In the first post on this blog we laid out the 38-per-year figure — and the part most parents find hardest to sit with is not the number itself, but the pattern behind it. The overwhelming majority of those deaths did not happen to families in crisis, or to parents who had been flagged by anyone as unsafe. They happened to working-memory failures: a parent running yesterday's correct mental model on today's disrupted morning. That framing is not comforting. But it points somewhere useful. If the failure mode is memory, then the intervention that fits it is also external memory — a reminder that lives outside the parent's head, arrives on a schedule, and does not require any thinking of its own to function. That is what this study is. For five consecutive weekdays we will send a single short text each morning to caregivers of children under five, asking one yes-or-no question about that day's drop-off and back-seat check. Each prompt takes seconds. None of them require an app, login, or any new habit to install. The pillbox did not invent medicine; it invented adherence. We are betting a daily text can do the same for the back-seat check.

Human memory is not a safety system — an external prompt on a disrupted morning will beat relying on muscle memory every time.

Why daily, and why SMS

We chose a daily cadence because the underlying failure mode is episodic. A biweekly reminder fades into wallpaper; a one-shot alert at the moment of driving lands too late. A daily morning text slots into a window where the caregiver has attention to spare — coffee, a moment before the car starts. We chose SMS because it is the lowest-common-denominator push channel in 2026: every phone has it, no install required, no special permissions, no opt-in to a second app the caregiver will forget exists. A pull design (an app the caregiver opens themselves) requires the very habit we are trying to build. A push design uses the phone network to do the reaching for us — exactly the kind of external memory aid a disrupted morning cannot generate on its own.

The 5-day protocol

Each weekday of the study we send the same kind of text, but tuned to the day. Day 1 is an orientation: a brief note explaining why we are doing this, a single one-tap confirmation of consent to receive the prompts, and an optional consent to share anonymized responses back with us. Day 2 is the simplest prompt we could design — "Did you check the back seat today?" — a single tap to answer yes or no, no typing, no thinking beyond the answer itself. Day 3 layers in a daycare check-in: did the provider confirm arrival today, and did silence from the provider get treated as out-of-the-ordinary rather than routine. Day 4 asks about the day's routine itself — was it the same as yesterday, or did something shift (different driver, different route, different drop-off order). Day 5 closes the study with a short reflection: which of the four working days felt easiest, which felt like work, and whether the participant would keep going on their own without our prompts. Every text takes fewer than fifteen seconds. None require the participant to start a new app or remember anything about us between prompts.

What we hope to learn

We will report four measurable outcomes. One: completion rate across the five days — what fraction of enrolled caregivers answer every prompt we send. Two: perceived intrusion, captured by a single self-report at the close of day five, since the entire design rests on the prompts being cheap. Three: behavioral change, captured by the day-four self-report about whether the participant added a back-seat cue or a daycare arrival confirmation in response to the prompts. Four: correlation between disrupted-routine days and self-reported back-seat adherence on the following morning's drop-off. None of these are lab-grade measurements. They are field signals — and signals are enough to decide whether a longer, more rigorous study deserves a year of work. We will publish all of the aggregate numbers on our Safety Stats page as the data comes in.

If you are a caregiver of a child under five, the study takes about fifteen seconds a day for one workweek, runs entirely over SMS, and asks nothing of you between prompts — join the forgetfulness SMS study →. As the data comes in, full results will land on our Safety Stats page →.

Every year in the United States, an average of 38 children die of heatstroke after being left in a hot car. That figure comes from the same NoHeatStroke.org dataset we cite across this site, and has held remarkably steady for two decades. Each one of those deaths was preventable. Most of them happened to loving, attentive parents — not neglectful ones. That is the part that should stop every parent reading this in their tracks.

A car interior can climb from 80°F to 125°F in 60 minutes — even with the windows cracked.

How fast a parked car heats up

Most parents imagine the danger arriving in slow motion — a car warming over a long afternoon while someone runs an errand. The measurements tell a different story. Researchers have documented car-interior temperatures at fixed intervals on an 80°F day, and the curve is unforgiving. At minute zero the cabin matches the outside air (80°F). Within ten minutes the temperature climbs roughly 20°F. By twenty minutes the cabin reaches about 110°F. By thirty minutes it is pushing 120°F. By the one-hour mark the interior regularly tops 125°F — even with the windows cracked an inch. And this is the 80°F scenario: on a 95°F day the interior exceeds 130°F in the same window. You don't die of heatstroke because someone forgot a child for hours. You die because they forgot for minutes.

Why children are most vulnerable

A child's body is not a scaled-down adult body, and the difference matters most in environments like a hot car. Children absorb environmental heat roughly three to five times faster than adults, because their surface-area-to-mass ratio is higher and their sweating response is less efficient. Their core temperature rises faster, and once it crosses 104°F, organ damage begins. By 107°F, heatstroke is no longer a warning sign but a medical emergency, with a mortality rate that climbs sharply for every minute of delay. There is also a brutally simple mechanical fact: a sleeping infant strapped into a rear-facing car seat cannot self-rescue. They cannot open a door, cannot roll down a window, cannot call for help — cannot do any of the things an overheated adult instinctively tries. By the time visible distress shows up, internal damage is already underway.

The caregiver scenarios that drive the risk

Public health researchers have spent decades cataloging the circumstances behind hot-car heatstroke deaths, and the cases cluster into a small set of recurring patterns. The largest category is a disrupted routine — a parent who normally drops a child at daycare, but today has the other parent driving, or is taking a different route to work, or has been asked to make an extra stop. The mental model the parent is running was correct yesterday, and the day before, and the day before that. Today it is wrong. The second pattern is the mistaken assumption that someone else has the child — a spouse, grandparent, ex-partner, or daycare provider — followed by the car heading to the office instead of the second drop-off. The third is the sleeping-infant scenario: a quiet rear-facing infant soothes into a deep sleep, the parent parks and walks away, and no cry triggers a check. The fourth is a change in custody or schedule — a child in the car on a non-standard day, with no rehearsal of the new routine. In every one of these cases, the parent is not careless. They are human, working from memory, under stress, on autopilot.

Habits that actually prevent it

Because the failure modes are predictable, the layered habits that prevent them can be too. The most consistently cited is the something-on-the-back-seat rule: a wallet, phone, or work bag placed on the rear seat on every single drive, so reaching the destination without retrieving it forces a back-seat check. Pair that with a daycare check-in protocol — an arrival text confirmed by the provider, with silence treated as out-of-the-ordinary rather than routine. Add a never-memory-alone rule: on any day the routine is even slightly off, add a physical cue that survives a distracted brain — a stuffed animal clipped to the seatbelt that moves to the front with the child, or a sticky note on the steering wheel. Finally, accept that human memory is not a safety system, and equip the vehicle with one: an automatic occupant-and-temperature detection system that escalates alerts (driver, emergency contacts, eventually 911) if the cabin becomes unsafe. Habits buy you margin. Hardware buys you a guarantee. Used together, hot-car heatstroke becomes a tragedy that simply does not happen.

If you made it this far, the 38 statistic is no longer abstract for you — and the habits above are exactly why we built HeatSentry. To see every number cited here (and the additional data we could not fit into a single post) laid out on a single timeline, see the full numbers on our Safety Facts page →. And if you are a parent, caregiver, or childcare provider willing to share how your own routines handle moments like the ones above, we would love to hear from you — join our forgetfulness research study → and we will be in touch.