“Small Data will control our businesses, homes, and even our cars and robots – all of which will define the world of tomorrow. Focusing on Small Data now could help us keep from being overwhelmed by it later and assure we will benefit from the massive rush to the Internet of Things,” says Rob Enderle, Principal Analyst, Enderle Group.

The IoT (Internet of Things) is one of the hottest buzzwords exciting the tech industry. This white paper describes the broader perspective of the IoT in light of the terms Cloud and Big Data, coins the term Small Data, describes the opportunities and the challenges – and explains the role of GreenPeak.

Big Data is everywhere now!

For the last few years, Big Data has been a big theme in the Information Technology industry. Maybe you think Big Data is far away, in clouds and server farms, well… it is not. I was recently researching buying a new car on one of the online internet websites. The next time I logged on to Facebook and it showed me an advertisement, for exactly the brand and the model I was considering buying. Surprised? Yes! Did I like it? Well, maybe initially not. It felt like someone was looking over my shoulder when I was visiting that site. But then again, what is preferable – a relevant advertisement or an irrelevant one? Big Data is here with us now, and it is here to stay, because it pays for the access to that tremendous source of information that we call “the internet”.

While doing some research lately I realized just how much faster, accessible and how much more accurate information is today, compared to 20 years ago. Just type: “how many homes are connected to the internet” in a search engine and you get a decent answer (600 million). This recently led to a funny dinner table discussion when my wife asked me a question. I pulled up my tablet to find her the answer, infringing on our rule-of-the-house of having “no computers at the dinner table”.

This in turn lead to a funny argument whether the question was about “wanting to really know the answer”, or “just making silly conversation”. Big Data has now even joined our dinner table. But on the positive side, isn’t it interesting how more informed and in-depth a discussion can become, when the information-gathering step becomes trivial.

Big Data, the Internet of People is big!

Big Data and the “terabytes” involved are stunning. Big Data is about the IoP (the Internet of People) that has created a lot of wealth for all of us. It also has helped to democratize society, helped people to socialize, or restored contact with friends and family we long lost contact with.

Critics may rightly express concern that Big Data is turning into a “Big-Brother-is-watching-you”. In a way there is a serious risk of the internet indeed invading our privacy and morphing transparency into control, and in that way destroying equal access to information for everyone. So, it will be important for law makers to keep taking the right steps to prevent Big Brother becoming an uninvited and unwelcome guest at our dinner table. It will become equally important that governments are establishing legal frameworks to ensure that they comply with that framework themselves as well.

Small Data, the Internet of Things will be massive!

Small Data is just in its early stage compared to what is coming. If Big Data is “big”, the emerging Small Data world will be massive and many times larger in total numbers and complexity. Many are predicting that Small Data is the real next big opportunity that will fuel the Information Technology industry for the coming two decades.

Small Data is generated by the IoT (Internet of Things) and are the trillions and zillions of small bits and pieces of information that will be coming from everywhere and continuously. Sensors will extract Small Data from every environment – every location – whether or not there are people in residence or not. Small Data will be processed by sophisticated algorithms; combined, analyzed, interpreted, and Small Data will float back to the controllers in that same environment fulfilling any conclusion that was determined by the algorithms.

Sentrollers (sensor, actuators and controllers) will be the devices that monitor and control our environment, our homes, our cars, our offices, our cities; sentrollers will be there by the billions, continuously collecting, generating and distributing “petabytes” of information. The amounts of data generated by the Internet of things will dwarf that collected by our current Internet of People.

One concrete example of Small Data in action are our automobiles. A modern car is a multifunctional machine that, in addition to its primary function of getting us from one place to another, also handles navigation, entertainment and communication with the external world. Essential is that all of these functions can be controlled by the driver from the driver seat with the dashboard. Modern cars have dozens of sentrollers informing us about/executing on the state of the engine, the state of the windows/doors, the tire pressure of the wheels, providing warning signals when a seat belt is not fastened, warnings that another vehicle is too close, or that a person is walking behind the car, unseen in the rear view mirror.

In a sense, the car is a “Small Data” machine, waiting for Google to discover and index it. In the same way as your computer device can provide a platform for targeted advertisements, so can your car. The recent acquisition by Google of Skybox Imaging allows Google an easy way to add your car to its database of smart objects.

Going from today’s Small Data in the smart car to Small Data in the smart home is just a small step. The turn-over of homes is a little bit slower than the turn-over of cars, so compared to cars, our homes today are still somewhat primitive. Here is one good example. Have you ever wondered why your car has central door-locking, but our homes have not? Everyone agrees with the fact that central door locking in the home may be even more useful than in a car, where the doors are very close to each other. So why are our houses still so stupid when our cars are so smart?

The smart home is around the corner, and in particular, cable and telecom operators have recognized the opportunity to sell to their subscribers not only TV or internet services, but also smart home services, the so-called Fifth Play services.

Small Data – gathered from all aspects of your life – not just your web search, marketing and sales activities, but everything – will help advertisers better target you and your family with advertising and marketing pitches that are customized for your lifestyle. Even knowing what your house is doing when you are not in it, will help advertisers become more efficient.

Small Data, massive business

Earlier this year (2014), Small Data arrived with a bang when Google acquired the company Nest that sells sophisticated thermostats and smoke detectors. This acquisition was not because Google all of a sudden wanted to compete with companies such as Honeywell or Bosch. It is about gathering Small Data in the Home. It demonstrates that in our evolving and evermore connected world, there is not a single place that will not be touched by innovation and changing business models. The classic thermostat is a typical dumb sentroller: a combination of a sensor, measuring the temperature, with a controller, allowing a person to set the temperature to a desired value, plus an actuator, a mechanism behind it to turn on or off the heater (or air-conditioner). That thermostat was not connected to the internet. The Nest thermostat, and other gizmos like it, are a part of the home data collection network – a true sentroller – with sensors, actuators and controllers all rolled into one device that has its intelligence shared in the cloud, Small Data.

So, what will the thermostat of the future look like? First of all, it will be connected to the internet, so it “knows” the outside temperature, as well as how the outside temperature will develop in the coming hours. It can calculate what impact external temperatures throughout the day will have on the internal climate of the home. It also will “know” whether people are at home, and where they are in the home, so it can direct which rooms need to be heated (or cooled). On top of that, it will “know” when people usually are at home or when they will come home, or what time they usually go to bed, based on the experience of previous days and weeks. Combining all this “knowledge”, this thermostat will help people to save energy and to reduce their energy bill, while making living more comfortable. So, with all this Small Data, the good-old, simple thermostat has evolved into a life style monitoring device.

Interesting right? However it is just the beginning. This is about the big business of Small Data collection and how it will provide the service provider of the smart thermostat service to understand their subscriber’s life style and will offer further help to target the right advertising to the right family.

But there is more. Understanding lifestyle enables recognizing exception situations, and creating alerts if these exception situations are undesired. Elderly people living at home longer will appreciate if they know that their family members will be warned if something unusual happens. An important concern for elderly people living at home (and for their children!) is the knowledge that if something happens and if help is required, that this will not go unnoticed. Of course, this status reporting can help more than just elderly people. For example, young parents can easily keep on monitoring the breathing of their newborns, and will be alerted if something out of the expected happens.

The Frog in the Pan

When you throw a frog into a pan with hot water it will immediately jump out of it – it can sense sudden uncomfortable warm temperature. But when you manage to get a frog comfortably into a pan with cold water and you start heating the pan, the frog will die from overheating, because it cannot sense slow changes in the environmental temperature.

In this sense people are like frogs. People can slowly gain weight over longer periods in time without noticing the carrying around of all that extra weight. By the time they notice it, the weight is incredibly difficult to get rid of. In addition, other body measures can quietly deteriorate – like blood pressure, cholesterol levels, interrupted sleeping patterns, etc. and can be silent indicators of needing to take action. If we would just be able to monitor our life style over longer periods of time, we can be warned, and we can be provided the right guidance.

Can you imagine how valuable this information will be for the marketers of diets, food supplements, exercise gear, the medical establishment, as well as for individual health providers who would really like to know what their patients are doing when they are not in the office?

There is a lot of data that can be collected, analyzed and acted upon. The challenge for Small Data will be how to reliably and securely collect this data in an unobtrusive and cost effective way. This is the major challenge that the industry is solving at this moment.

How Large is Small Data?

We all witnessed the amazing evolution and success of WiFi, growing from an interesting, obscure technology, struggling to find market acceptance for almost a decade, but then suddenly turning into a key enabler for Big Data: allowing people to share information anywhere and anytime, from our offices to our dinner tables, from industrial applications into airplanes, trains, cars. About 15 years ago WiFi needed to be sold as an add-on accessory. Today WiFi is included in every Big Data device (computers, laptops, tablets, smart phones, TV’s, game stations, etc.) for receiving and sharing information. Every home that is connected to the internet (600M today) has about 10 Big Data devices – leading to a market of 6 billion devices.

In contrast, our homes could easily accommodate 100 or more Small Data devices, ranging from temperature devices, to motion sensors on doors and window, lights and locks, etc. This means that for the coming years there is a potential hardware market of 60 billion Small Data devices. Every device will probably cost less than $10, which means that over time every home will become “smart” for less than $1,000, establishing a $600B market for just the hardware sentroller devices. For many, these devices will probably be subsidized by subscriptions – comparable with today’s smart phone services. Worldwide, operators are currently testing the various business models to offer these Smart Home services.

Key Challenges for Small Data

The concept of the smart home has existed for years. The technology for the smart home has also existed for years as well. But it never came together in a practical, affordable manner. There was no connecting technology – there was no connecting intelligence.

The economics for the smart home are now being worked out and are quite challenging. From a distribution model perspective, there are already several contenders providing different go-to-market models. As expected, companies like Apple, Google, Samsung – all with a strong smart phone perspective, are developing market models. In comparison, cable and telecom operators like Comcast, Liberty Global, and AT&T are extending their subscriber services with smart home service like security, energy management or assisted living. There are the retailers like Staples, Home Depot, Walmart, Amazon, etc., positioning themselves for the smart home market as well. There is a great deal of market interest and competition trying to establish which business and technology model will provide the best service at the best price. It is very exciting.

The smart home should add simplicity instead of complexity, so it needs to be based on standards, a simple architecture. It needs to be private, secure and easy to install and maintain and totally non-obtrusive. In the same way that the smart systems in a car are not obtrusive, automatically turning on the headlights when it is gets dark or when driving into a tunnel.

Also, the economics need to work. Companies that successfully figure out how to extract life style monitoring from Small Data collections and convert that data into non-obtrusive targeted advertising supported models, will be the ones that succeed in the market. Key to this is the economics involved for the cost of developing these products and then the necessary economy of scale to drive down these costs. Essential for these economies of scale are worldwide uniform standardization at all levels, from how the data is collected at the lowest levels, how that data is distributed over the internet, how and in what format the data is stored, and how it is made available to application builders developing the lifestyle monitoring applications.

Standardization is a challenging process although good progress has been made. Small Data at the sentroller level will be collected and transported via ZigBee (Low-power WiFi) for the smart home and with Bluetooth Smart (Bluetooth Low Energy) for the wearables. Long battery life at the end nodes is critical and these communication technologies have been developed especially for long battery life.

At the networking level, IPv6 was developed to distribute Small Data over the worldwide web, and although IPv6 will be a challenge for the cost and the battery life of the end nodes, solutions will be developed to resolve these issues. The advantage of ZigBee, Bluetooth and IPv6 are that they are open worldwide standards, enabling worldwide uniform product implementation, without single source dependency.

Another big challenge is that market needs to become comfortable with privacy and security issues before fully embracing Small Data as part of life style monitoring. Security is clearly a never ending rat race between users and hackers, where every improvement in security is followed by a next level of sophistication of hackers. In general the technologists have realized that there is no ultimate security yet, and possibly never will be, so security will also need to include recognizing whether and how systems have been hacked and can be secured again. This also fully applies to Small Data.

Big Industry impact of Small Data

New concepts and new horizons usually send shockwaves through the Information Technology industry. The arrival of Big Data has led to many companies dying (who remembers 3Com), merging into new companies (for instance Lucent Alcatel) trying to reinvent themselves (for instance Microsoft) and/or the establishing of new companies (like Google).

It is no surprise then that the arrival of Small Data is both an opportunity as well as a threat and is accelerating strategic activity in many large companies. At the same time it would be no surprise if Small Data will generate a set of new companies, with Nest as just one early example.

The impact of Small Data on the software and service industry will be in the area of pattern recognition, artificial intelligence, and from there, a step up to predicting behavior and sending alerts for recognition of exception situations. At the same time, more demands will be placed on security and protecting privacy, while maintaining ease of installation, and great out-of-the-box experiences – usually quite conflicting demands.

The impact on the hardware industry of Small Data may even be bigger. Both the personal wearable devices as well as the smart home devices required for providing Small Data information will become a deluge. The biggest impact however will be on the semiconductor industry that can look forward to a new period of growth in particular around low power embedded microcontrollers and radio communication devices as well as sensor technology (in particular MEMS). Each sentroller exists as one or multiple combinations of these technologies. Over the next decade, the demand for these devices will grow exponentially, very comparable with the growth of the laptop industry 15 years ago, or the smart phone industry the last few years.

Small Data System Architecture

The picture below is a schematic overview of the basic architecture of Small Data.

The key devices in the Small Data Architecture are (1) the Internet Gateway, connecting the smart home to the cloud, where the Data Analysis and Action System is running and (2) the Smart Phone, functioning as a dashboard and controlling the interaction with Small Data, like threshold settings and receiving alerts. The Internet Gateway as well as the Smart Phone are both connected to the cloud, via cable (DSL) or wireless: cellular (2G, 3G, LTE, etc.) or WiFi via a hotspot or again a gateway.

Within the Smart Home network, the Internet Gateway connects all the Smart Home sentrollers via ZigBee to the cloud for data collection and action. The Smart phone can connect to the wearable sentrollers via Bluetooth that are going around with a person, whether at home or on the road.

Interestingly one can foresee in the Smart Home so-called dual-mode sentrollers: sentrollers that support both ZigBee and Bluetooth – think for instance about a door-lock. If the internet gateway, or the internet is down, it would still be nice to be able to open or close the door with the Smart Phone.

This is a basic picture showing the fundamentals of Small Data systems. The big advantage of this architecture is that it seamlessly fits in the overall internet structure as we know it today and that it uses open worldwide standards (ZigBee, Bluetooth and IPv6). Therefore it can be seamlessly extended as well. For example, many smart home systems today have proprietary dashboards or displays. These can be completely eliminated as essentially with the right security in place, any Smart Phone, tablet or PC can run cloud based dashboard applications that can monitor and control a wide range of sentroller applications. Also Big Data and Small Data applications can be mixed without changes in the architecture, like combining a video security camera (from the Big Data world) with a motion sensor (from the Small Data world) – the IoP and the IoT growing into the IoE (the Internet of Everything).

“The Internet of Everything is a dynamic environment where the synergies between Big Data and Small Data are realized, enabling connectivity and embedded intelligence for an extremely broad range of devices, many of which are not connected today. The IoE will be populated by devices that collect Small Data from their environment and then use analysis of that data (accessed via the cloud and analyzed using “Big Data” techniques) to provide unique value propositions and create information systems which are greater than the sum of their parts. It is an interesting concept and represents essentially a new way of doing business and a new way of offering services in a booming market.” Says Dr. Jagdish Rebello, Senior Director Cloud and Computer Electronics, IHS.

Small Data Eco-systems

It is clear that Small Data systems will require a lot of cooperation between contributing parties to make it seamlessly work together. All the technology pieces are currently available in one or another format, but the devil is in the details to make things work cohesively and be economically efficient.

Start with the end-customer, the consumer. The consumer will need to buy sentroller end products from the cable or telecom operator and/or the retailer, who will also offer a subscription for instance for the life style monitoring and for instance home security or energy management systems. When products have enough standard interfaces, the retailer can also sell products on behalf of the operator.

The operators and the retailers are packaging and selling this product/service combination by combining different products into single offerings. They can leverage their position in the value chain in two ways: from a brand perspective (consumers know and trust large brands), as well as from a customer touch point perspective. In particular operators have an advantage here, because they have already the recurring touch points with theirs subscribers in place – for instance via monthly billing systems. They have also the call support centers up and running, for which another service is just incremental business.

It is also possible that an operator will choose to use a no-brand hosting service instead of running the system themselves. In that way a large brand retailer can be a service provider, using such a no-brand hosting service. The hardware products (sentrollers and the gateway) are developed based on the requirements of the operator or retailer and are manufactured by low cost/high volume manufacturers.

Software is an important key to the system. There is the data collection piece essentially connecting the sentrollers with a database that is stored in the cloud. There is the data analysis and action system that runs in the cloud and includes the engine for interpreting the data and generating actions, whether these are sending signals back to the sentrollers to make adjustments or alerts to a smart phone. In the smart phone there needs to be a dashboard application that can also handle alerts.

In addition to this software, there is also an administration system required, connecting phone apps with subscribers, and putting the necessary functions and architectures for privacy and security in place.

Advertising can also cushion the cost and speed the acceptance of Small Data devices in your home and car. Would you allow Google to track your movements – your family movements – within the home, and then sell that data, if your service provider offered you a substantial discount on the smart home services? Advertising essentially supports the current Internet of People – doesn’t it make sense to allow advertising to also support the Internet of Things?

The Role of GreenPeak Technologies

GreenPeak Technologies is a technology company focusing on Small Data application systems. Collecting, managing and extracting useful information from the massive volumes of Small Data is a tremendous technology challenge, and an opportunity for new, innovative companies.

Over the last few years, GreenPeak is one of the companies that has been driving the development and building of these Small Data systems through:

  • Developing and owning core technology pieces that collect Small Data from the end nodes in a cost efficient way.
  • Licensing-in certain technology components from third parties to complete Small Data Systems so they can be used by Operators and Retailers.
  • Developing strategic relationships with hardware and software providers to design and manufacture easily installed and maintained Small Data systems

End-to-end system understanding and integration capabilities are essential for bringing to market solutions that are customer friendly and economically feasible. There is a reason that Smart Homes have been promised for a long time, but still have to live up to those promises: Small Data has been looked at as a simple group of sensors, but in reality it is a system that stretches from these sensors into the Cloud and into Big Data.

GreenPeak Technologies’ strategic value is that it can deliver to the brands (Operators and Retailers) complete Small Data systems, including all the hardware and software components that will drive the next wave of Information Technology, the processing of Small Data in the Internet of Things and the building of the Internet of Everything.