In building PreSafe, we started with the assumption that locating the person who was wandering was the only concern, and we decided by build a GPS locator that had a much longer battery life than what we could find on the market. The reason for placing such an emphasis on battery life, was straightforward for us: the adult children of parents with Alzheimer's were not always living with that parent, and oftentimes they could visit only on weekends. This meant that the GPS locator we needed to build had to have at least seven days of battery life, compared to the one or two days that were say reported in consumer reviews of existing GPS locators.
The other key focus we had was on not using a wristband, or an amulet, because we learned from research at Assisted Care and Memory Care facilities that amulets were very quickly removed by the residents of professionally run facilities, and that wristbands were either quickly removed, or chafed at endlessly.
Design was also an important consideration. Studies done by AARP found that wearing or carrying devices that looked "medical" were resisted by everyone, because such designs tended to make the wearer feels stigmatized and shunned. Regardless of one's age and cognitive deficits, one always wanted to look at their best and most capable.
So, within those constraints, we proceeded to select a hardware chipset that could accommodate those goals, and focused on then keeping the costs down to the most affordable level we could. That meant, in that early stage, opting for a 2G network, a GSM modem because of its small size and low cost, and making the assumption that we could use the 2G/GSM combination with the T-Mobile network until 2017—when T-Mobile would begin deprecating that network (ATT already had started taking down their 2G network). That T-Mobile 2G network ran in the 1700 mhz range, which would become important for us as we began development.
We selected the chipset we would use, the accelerometers, and GPS chip, and the Modem and Microprocessor, and began to build "spikes" to prove out concept and the end to end functionality we needed.
It wasn't enough to know the location of someone, it was just as important to know whether or not they were in trouble
And, it was at about that time that we learned two new things: First, it wasn't enough to know the location of someone, it was just as important to know whether or not they were in trouble – which could be completely independent of location. Witness the recent tragedy from the recent Northeast blizzard, Juno, where an 83 year old man froze to death in his backyard during that storm. A GPS locator would have never shown that he was in trouble, because it would have looked like – on a map – he was safe at home (GPS locators won't reliably show when you are inside, or outside, your back door.
This meant further research into what the actual causes of injury were, related to wandering and Alzheimer's. While research was thin on the topic of what actually causes the serious injuries for people who have gone wandering, we found that there were three key areas of vulnerability:
(1) Injury from hypothermia and heat exhaustion/stroke,
(2) Traffic incidents, both while driving and walking, and
Falling was also found to be a significant risk factor, however it was independent of wandering behavior.
After much time, we determined that we could address two of the four causes of injury: Hypothermia/Heat Stroke, and Falling. We committed to using a temperature sensor to detect significant changes in the external environment, and an accelerometer to detect falling. Guarding against traffic danger and drowning risk would require a great deal more research, which would have to be delayed for the product's first release.
The second thing we learned was that using a high-frequency 1700mgz network was going to lead to occasional failures in getting a signal out from indoor spaces. This became clear to us when we researched the reviews of existing generic GPS locators for children, where the users complained that the system was not showing when they were in their school buildings or other indoor environments.
After much hand-wringing, and even more research, we determined that the best way to solved the problem getting access to a relatively low frequency network was to switch from the low-cost GSM modem to the more expensive, and large, CDMA modem that would run on the Verizon network.
While this solved the problem for the United States, it did not yet address the issue for the EU and Asia. Mostly, GSM was the standard on both those continents, and fortunately 2G networks would still be in use for many years to come, in the 900 mgz range.
This left only one more issue: how to get a GPS fix when the individual was within a windowless environment, where there was no way to get a satellite fix. For that issue, we needed to use trilateration, where the modem that pings the cell tower(s) can use information gathered from signal strength and relative position to approximate the location of the PreSafe device. Presently, that's our approach to getting an approximate fix on the device location, enabled by the location-based servers provided by Verizon and our chip manufacturers as well. However, its approximation has an standard deviation that approaches 100 meters, so we'll be researching way to significantly improve on that approximation in the near future.
For the present, then, we've been able to address to a meaningful degree a variety of hazards that befall individuals living with Alzheimer's: becoming lost or wandering both indoors and outdoors, injury from extreme weather, and rescue from falling. And, we've addressed the issue of long battery life, extending the battery life of PreSafe to an order of magnitude greater than generic locators through heuristic algorithms.
Our goal is to make the persons living with Alzheimer's much safer than they had been, previously, and to allow their caregivers the opportunity to breathe more easily when there is a job dominated by stress and pressure.