Page 1
Essence Care@Home

Senior monitoring – security flipped on its head

Let’s say you want to set up a security system for your home. In days past that would mean hiring a security company that would install sensors around your house, connected to a keypad, and linked via a phone line (usually a dedicated line you would need to add) to a monitoring company. Those sensors would probably have required wired connections to a central hub installed in a closet somewhere, or in a drop ceiling, requiring lots of installation work, cutting holes in walls, and cleaning up the mess afterwards with spackle and paint. Once installed, this system would only work if you paid the company that installed it a monthly fee.

DIY Security Systems

In recent years a new category of security systems have emerged, so-called DIY security systems. These systems are designed to be installed by the homeowner, and in general do not require ripping up walls to install them. Some systems allow you to monitor your house yourself, and some include monitoring for a fee, similar to the older systems. One good example of this type of system is SimpliSafe, which sends you a kit including various sensors to install yourself, and then provides a traditional monitoring service. In general these new systems use wireless connections to connect to a central hub, and each device therefore only needs power. This means you can simply plug in a sensor and connect it to your hub, wherever you have an electrical outlet. Smaller sensors can work with just batteries and can run for up to a year before needing the batteries to be swapped or charged.

Camera-based Security Systems

Blink system with 3 Cameras
Blink system with 3 Cameras

More recently, systems have come out where cameras, the most power-hungry of security sensors, have been able to be run on batteries as well. Whole security systems have been developed based on just using these wireless cameras with batteries. Examples include Netgear’s Arlo system, as well as Blink, Homeboy and Canary.

All of this innovation is great, but runs into a problem when one wants to set up a system to help their elderly relative. Much of the technology is the same, but usage is flipped on its head. You don’t want a system that is triggered every time it senses movement, but one that recognizes when there is a lack of movement. While a standard security system can be dumb in that any movement is considered bad, a system set up to monitor the elderly should be smart and learn patterns of movement, only notifying when there is a change in the pattern.

Some security system are cognizant of privacy concerns, offering physical shutters to block cameras when a system is disarmed, and insures all video streams are encrypted. When monitoring the elderly, however, your system is never really disarmed, it’s just in a different mode. If your security system is camera-based and gets its motion-detection capabilities from the camera, then closing the shutter means your system cannot operate at all.

Of course, having a camera on all the time has major privacy concerns, and your elderly relative may, understandably, not want a cameras watching them at all times, even if its just for motion sensing. All of these new camera-based systems sadly cannot be used for the elderly, or at least the system cannot be solely based on cameras. This means another type of solution is required.

Existing Security Systems Adapted for the Elderly Wellness Notification Wellness Notification

Some existing security system companies offer versions adapted for the elderly., for example, offers a service called Wellness that monitors seniors in their home, recognizing patterns, and even sensing how long someone stays in bed or in a favorite chair. This system is integrated into their monitoring service, and can alert a family member if their relative leaves the house at a strange time, or doesn’t get out of bed all day, or doesn’t open the fridge.

Essence [email protected]

Essence Care@Home
Essence [email protected]

Israeli company Essence, which sells panic-button systems for use in elderly homes, has come out with a more comprehensive system called [email protected] that uses a combination of motion sensors, door sensors, cameras, etc. as well as a traditional panic button, in combination with a service to look for patterns and notify family and care providers accordingly. This system is sold to monitoring companies that install the system in people’s homes and sell a monitoring service. Silver Mother

Silver Mother Sensor on Medicine Bottle
Silver Mother Sensor on Medicine Bottle

One interesting solution is from French company, which offers a product called Silver Mother. Silver Mother works with a central hub with small wireless motion sensors. These sensors can be used many ways. For example, place one on the refrigerator door to sense how often the fridge is opened. Place one on a mattress to see how often and how long a person is in bed. Put one on a medicine bottle to know that someone has taken their medicine (or rather at least one can determine if they have not taken their medicine if the bottle never moves). Another advantage of the Silver Mother system is there are no monthly fees. It is for the most part a self-contained system, although it can work with Nest and IFTTT.

DIY Options for Senior Monitoring

If you wanted to build your own senior-focused security/monitoring system, there are a few options, although none that are perfect. One approach is to use a multi-purpose hub like Samsung’s SmartThings Hub, or Wink’s Hub. Both products connect to devices from many different manufacturers, using multiple home-automation protocols, such as Z-Wave and Zigbee, as well as WiFi and Bluetooth. Wink additionally works with some manufacturer-specific protocols like ones from Kidde and Lutron.

Protocol Problems

The use of popular home automation protocols like Z-Wave and Zigbee is key, as it allows many devices from many manufacturers to be used together. Unfortunately, that’s only partly true. Both protocols have their problems.

Z-Wave uses different frequencies in different countries, and there is no such thing as a hub that can handle more than one frequency. If you’re in the US and you buy a US hub, you need to buy sensors that are intended for the US market. There are more than a dozen different frequencies used around the world. The US is different than Europe, which is different than Australia, which is different than Japan, which is different from China, etc. Hong Kong shares one frequency used in the US, but not a second one. For a large multinational company that can manufacture dozens of versions of their products, this is okay. For a small company looking to break into the home automation market, this is a major problem. If you move around, you also may not be able to take your equipment with you.

Zigbee can for the most part be used on a single frequency (although it does support using similar frequencies to Z-Wave in some cases) worldwide because it works in the same frequency as WiFi and Bluetooth (2.4 GHz), although it uses different profiles for different devices, and devices designed for one profile will not work if the device its connecting to uses a different profile. For example, the Samsung SmartThings Hub supports the Zigbee Home Automation profile, but not the Zigbee Smart Energy profile. Add to this that the signal strength allowed in different countries can be different (the US allows almost twice the signal strength of Zigbee devices as is allowed in Europe, so if you buy a Zigbee device in the US it’s probably illegal to use it in Europe).

Programming a DIY System

Putting aside these protocol problems, we run into another problem. You need to program the system to notify you based on the sensors in ways that make sense for seniors. The non-DIY system have built that intelligence into their system. If you build something yourself, you need to figure out a way to create a similar intelligent method for notifying you. Some types of notifications are easier than others. For example, if the door to the outside opens between 11pm and 6am, send a notification. If the resident hasn’t gotten out of bed by 11am, send a notification. If the system sees no movement for over an hour during the day, send a notification. These are simple rules that could be enhance by using advanced pattern recognition, but still will work for the most part. A more sophisticated system would know when the resident gets up every day (i.e. between 8am and 9:30am) and would know if something is out of the ordinary (the resident is not yet up by 10am), but if you watched the data for a few weeks, you could probably just set the notification for 10am and get the same thing.

Both Wink and SmartThings can be programmed using IFTTT. This allows some level of interoperability between them, as well as with other systems that don’t support the same protocols, but do support IFTTT. IFTTT stands for IF This Then That, and is a simple rules-based system for telling Internet-connected services to trigger actions based on certain conditions. These sets of rules are called Recipes. For example, you could set up a recipe to send you an SMS whenever a specific stock went over a certain price, or you could get an e-mail whenever a specific product showed up on eBay. In Home Automation scenarios, you could have a recipe that whenever someone passes a motion sensor outside your door, the light outside is turned on and a beep is sounded inside. These can be very powerful, but are limited in intelligence. You could use an IFTTT recipe to turn on a light through your Wink Hub when the sensor in a Silver Mother system detects the resident gets out of bed (although you would need both the Wink hub and the Silver Mother hub).

Samsung gets Groovy

Samsung, in addition to supporting IFTTT, also supports creating programs using Groovy, a language developed by the Apache Foundation which works in the Java Platform. This means it can use Java libraries, and is run through the JVM. The technical details don’t matter too much, but in short it means that you have a lot more control in building intelligent applications using SmartThings than you do with Wink.

As an example, here’s a program I found posted in the SmartThings Community that triggers a notification if there is no motion between a specific start time and an end time:

 *  Notify if no motion
 *  Copyright 2015 Bruce Ravenel
 *  Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except
 *  in compliance with the License. You may obtain a copy of the License at:
 *  Unless required by applicable law or agreed to in writing, software distributed under the License is distributed
 *  on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License
 *  for the specific language governing permissions and limitations under the License.
    name: "Notify if no motion",
    namespace: "bravenel",
    author: "Bruce Ravenel",
    description: "Notify if no motion during some time period",
    category: "Family",
    iconUrl: "",
    iconX2Url: "[email protected]",
    iconX3Url: "[email protected]")

preferences {
	section("Which motion sensors?") {
		input "motions", "capability.motionSensor", title: "Motion(s):", required: true, multiple: true
	section("Text me at (optional, sends a push notification if not specified)...") {
		input "phone", "phone", title: "Phone number?", required: false
    section("Unless this presence sensor is not present...") {
	    input "presence1", "capability.presenceSensor", title: "Which presence?", required: false, multiple: false

        section("If no motion between these times...") {
		input "starting", "time", title: "Starting", required: false
		input "ending", "time", title: "Ending", required: false

def installed() {

def updated() {

def initialize() {
	subscribe(motions, "", motionHandler)
        schedule(starting, startingTime)
        schedule(ending, endingTime)

def motionHandler(evt) {
	state.noMotion = false

def startingTime() {
	state.noMotion = true

def endingTime() {
    if(presence1) if(presence1.currentPresence == "not present") return
	if(state.noMotion) {
    	    def msg = "There has been no motion since $starting"
	    if (phone) sendSms(phone, msg)
            else sendNotification(msg)

I don’t mean to show this as an example of a complex program, but rather just what the code looks like for a simple program. Samsung has a Marketplace within their phone app that allows you to download many applications (dubbed SmartApps in keeping with their naming pattern), including a whole category called Elder Care. I wasn’t able to find a list of the SmartApps in that category online, and I can’t look at them in their app because you need to connect it to a Hub before you can view the Marketplace.

The average person isn’t going to develop machine-learning algorithms in Groovy to learn the daily activity patterns of their elderly relative. Hopefully enough relevant SmartApps exist in Samsung’s Marketplace that they wouldn’t have to do it themselves.

Where does that leave us?

It would appear that we have either pre-packaged systems that have intelligent activity monitoring which requires monthly payments, or DIY systems that are a bit touch-and-go considering you need to buy lots of sensors and hook them into some existing applications that may do what you want, or you may need to write custom applications to get there, and none of those are likely to have the full level of intelligence needed to prevent frequent false positive notifications. There are some good DIY systems like Wink and SmartThings, and even senior-focused systems like Silver Mother, but they’re not set-it-and-forget-it easy to configure. We’re at the point where system like Blink have pretty much brought that easy of installation to the home security market, but the senior market is still catching up. It should be interesting to see how this niche makes up the ground in the next few years.

The end
Asus ZenWatch and Sony SmartWatch 3

On the brink of an Apple wearable

Moto 360 and LG G Watch R
Moto 360 and LG G Watch R

Back in April I wrote a post titled One Wearable to Rule Them All? Not likely. predicting where I thought the wearable market was heading. My basic idea was that there are essentially three types of wearables – the all-in-one, the single-purpose, and the multi-position.

The all-in-one is where the smart watch category is headed. These wearables will look like watches, have sensors to track many different health issues (steps, motion, heart rate, temperature, blood oxygen, etc.), and be able to assist in authentication and commerce. This is in addition to an array of information-realted apps, and an optional link to your smart phone.

The single-purpose is best represented by fitness bands that track health data or bands like the nymi that can be used for security applications. These bands are almost always worn on the wrist, and are focused on a single application. These wearables can be wholly replaced by an all-in-one wearable.

The multi-position is a variation on the single-purpose, is usually focused on health, but can be worn in several positions, such as your wrist, your ankle, etc. and be used for many sports such as running, biking, swimming, etc. The best example of this is the Moov, although wearables like the Sony Core and the Fitbit Force are moving in a similar direction. These wearables can’t be easily replaced by a wrist-worn wearable such as a smart watch, although some aspects of what they might be done better by an all-in-one.

In April the first Android Wear watches had already been announced, but had not all shipped yet. The flagship of the early announcement was clearly the Moto 360, Motorola’s round-screen smart watch. That only shipped in the past week or so, and immediately sold out online. Selling out sounds great, but I suspect it is more about constrained manufacturing than success in selling their wearable.

In the intervening months other Android Wear watches have been announced, several in the past couple of weeks in the lead-up to Apple’s presumed announcement of their entry to the wearable space (in just a couple of hours as I write this).

Asus ZenWatch and Sony SmartWatch 3
Asus ZenWatch and Sony SmartWatch 3

Sony released the SmartWatch 3, now based on Android Wear. Samsung seems to have rushed the announcement of their Gear S watch, which has a curved screen and health functions, but has no release date in the US any time soon. Asus announced their ZenWatch which has one of the most refined looks of any of the Android Wear watches. Perhaps the most interesting is the LG G Watch R, which competes directly with the Moto 360 as a round-faced watch. The LG G Watch R is a major improvement on LG’s earlier announced G Watch from April, with the round face, metal bezel, water resistance, health tracking, etc.

Interestingly, the G Watch R is the first Android Wear watch to have a completely round face, as the Moto 360 has a small portion of the bottom of the round face that has no screen. Motorola has argued that in order to have a fully round face, you need to add a bezel around the edge. LG indeed did need a bezel, but turned lemons into lemonade and made the bezel similar to those on metal sports watches, giving it a traditional look that no other smart watch currently matches.

Google has been hard at work trying to add features to Android Wear, the operating system all of these watches are based on (except the Samsung Gear S, which runs Tizen). Future releases of the OS will support connecting to other Bluetooth devices such as headphones, support GPS (although I don’t think any current hardware has a GPS chip), and will amusingly support easier-to-build watch faces. Even so, there are some gaping holes across the Android Wear ecosystem such as a unified health tracking system and commerce/authentication features.

The Elephant in the Room

Elephant in the Room, from the Banksy Barely Legal art exhibit in 2006

Apple of course is strongly suspected to be launching its entry into the wearable space, and even when leaks surround the next iPhone have been more numerous than for any Apple product I can remember, there has been almost nothing leaked concerning Apple’s wearable, generally called the iWatch, but for which even the name remains unknown.

It is suspected, however, that Apple will tick off almost all, if not all, of the features I predicted back in April in their wearable. In particular, it is expected that the Apple wearable will support commerce applications, and will introduce NFC-support in order to do so. Apple already announce their HealthKit framework for supporting health sensors and applications, and it is expected that the Apple device will contain many sensors to drive HealthKit applications. It goes without saying that Apple has taken on the issue of the user wanting the wear their device, and if LG currently holds the title for most refined-looking smart watch briefly, that is likely to be lost quite soon.

So here is us, at the raggedy edge, waiting for Apple to announce their entry. If they support gps, payments, health and authentication, will all current smart watches be immediately obsolete? Perhaps not, but the elephant in the room will be stepping on a lot of current contenders for the smart watch throne.

The end

Samsung looking to Israeli tech to compete with Apple’s iBeacon

A couple of months ago I tweeted about an Israeli startup named ShopCloud that was showing off some fairly amazing retail technology:

The interesting thing to me was that it enabled much of the functionality of Apple’s iBeacon technology, without the need for physical iBeacons to be in place.

It seems this fact wasn’t lost on Samsung, which is now rumored to be trying to buy ShopCloud for about $80-90M. This was originally reported by the Israeli tech blog Geektime, and followed up by Israeli business news site Globes.

ShopCloud INSIDE
Navigating using ShopCloud’s INSIDE

This is particularly relevant if you look back at my post Who do you trust with your identity? which among other things looks at how Apple is using iBeacons to position itself as the preferred partner for mobile payments in the future. It’s a smart strategy, giving the retailers powerful technology to engage their customers, while at the same time giving Apple access to those same customers. Google and various other companies have tried to use NFC to similar effect, but these efforts have largely failed.

ShopCloud could allow Samsung, or whomever ends up purchasing it, an end-run around iBeacons and NFC. ShopCloud’s INSIDE technology allows malls and store to map everything to a 1-meter accuracy, and let the user navigate through the store or mall easily. Just like iBeacons, the app could make offers based on location. If the technology works as advertised, then the costs for deployment are significantly less expensive than deploying iBeacons, with many of the same benefits. This gives the owner of the technology a leg up on Apple’s iBeacon, and similar access to retailers.

I think Samsung would be very smart to buy ShopCloud, as the competition for who will control mobile payments is definitely heating up and in the next few years we’re going to see a handful of companies controlling those payments. Samsung definitely wants to be one of those companies. If a bidding war erupts over the company, I think you’ll see some of the other companies interesting in mobile payments also getting involved.

The end