We are living in exciting times – IoT is here, Analytics is here, and the world is full of smart gadgets, smart behaviour, etc. The terms are brandied about so much, at times bordering on hype, that there is confusion among the consumers. The billboards and advertisements all around him/her are screaming – “Smart phones”, “Smart TV”, “Smart homes”, “Smart energy”, “Smart schools”, “Smart plugs”, “Smart manufacturing”, “Smart cars”, “Smart healthcare”, “Smart cities”, etc. One wonders what this smartness is about. What is the difference between a Smart item and a non-smart item? Is it that something is either smart or non-smart? Scratching the surface reveals that it is not binary (smart or not smart) but a maze of continuum in smartness. In this post, I try to explain this maze of continuum.
The smartness in Smart systems can range right from being totally non-smart to being like Mr Spock in Star Wars. Smartness is at the overlapping region of a few aspects; some of them are Data-availability, Analytics-capability and UX-design. Let us dig deeper into each of these aspects.
We are flooded with data – too much data – thanks to the thousands of sensors everywhere. There are sensors in every equipment and gadget around us, even those which we considered inconceivable for long. Who would have thought a simple light bulb would be a source of data and not just light? And a traffic signal would also be a source data too? These serve various needs for various end-users in various situations. A few examples follow.
- Smart phones. In the initial days, Smart phones had very few sensors – like touch and GPS. The newer generations of Smart phones have dozens of sensors – gyro, accelerometers, etc. These generate so much data constantly that many applications are dependent on them; for example, location-based services, determining geometry of environment, movement sensing, etc.
- Smart homes. Modern homes have so many gadgets, appliances, equipment, etc. which generate such huge data that homes built even ten years ago seem like stone-age constructions. For example, there is a lot of focus on energy management or power saving. These applications depend heavily on the data generated by sensors in systems like the HVAC systems.
- Smart cars. Modern cars have a hundred small computers under the hood and have many sensors like some for engine efficiency, some for driver assistance, some for passenger infotainment, some for security, etc. It is a case of data-abundancy, not just availability.
- Smart manufacturing. Factories usually have thousands of sophisticated equipment from different vendors, all designed to work in harmony to produce the desired goods efficiently. Goods are indeed produced, but the efficiency is as critical as the goods themselves. The various equipment, apart from doing their jobs in the process of manufacturing, also generate lots of data. These may be temperature data, pressure data, density data, so many electrical parameters, etc.
Analytics is the brain behind Smartness. Analytics of data depends on what our interest is in various scenarios. In the early days, simple analytics was limited to taking binary decisions (yes/no, left/right, etc.) based simply on reception of data elements. This was sufficient for many situations like detecting the presence of an object by an RFID signal. Over the years, more challenging needs have given rise to more sophisticated analytics techniques. These analytics engines receive many inputs, execute multiple stages of algorithms, and provide rich results like trends, probabilities, recommendations, etc. A few examples are in order.
- Smart home. There is a lot of data constantly being generated by sensors in the appliances, HVAC systems, AV systems, Security systems, etc. Typical interests for Analytics range from analysing the data to minimize energy consumption, to detect intruders, to auto-tune the AV systems depending on the presence of various home members, etc.
- Smart cars. The key interests in this domain are to minimize driver distractions, to increase safety, to enhance drive experience among passengers, etc.
- Smart manufacturing. The data from all the equipment in the factory can be analysed to improve efficiency (of the full factory), maintainability, etc. The analytics usually cover efficiency, quality, performance, health/life of various components, etc.
- Smart sales. When big organizations produce goods in huge volumes, and marketing and sales forces need to be organized and run smoothly like a well-oiled machine. This machine generates lots of data regularly, like market data, sales data, data about partners, data about which practises yield what results, even data about which sales personnel are more successful in what circumstances, etc. This data can be analysed to increase sales and revenues of the organizations.
Experience reveals that UX-design is not just simple menu and buttons to click. This is a discipline so complex that it is a career choice for a growing number of professionals. There is so much to convey to the end-user, interactively, at the same time taking care not to overwhelm him/her with a clutter. Also, what use is the availability of so much data, so much analytics, if it is not fed back for use by the end-user? This step is the fulfilment of the purpose, the closure of the loop, and is often the deciding factor when people decide to purchase or appreciate products. Below are some examples to convey the criticality of UX-design.
- Smart phones. The Smart phones of today have ever-increasing capabilities related to display, resolution, graphics, memory, performance, software applications, etc. The results of analytics – be it text, data, excel, images, recommendations, probabilities, etc. – can be presented to the user. Interactivity, customization, personalization, context-awareness, localization, etc. are critical and are so common place that most users take them for granted. All this has resulted in most people treating their phones as their most dependable life companions!
- Smart cars. Nowadays, cars have lot more than an engine, wheels, body, air-conditioning, etc. There is a lot of electronics, processing and communication capabilities. The primary users of cars – drivers, passengers, etc. look for a superior drive-experience, infotainment, personalization, etc. by utilizing all the analytics, etc. and closing loops with the users through the facilities available in the cars themselves. These can happen through audio feedback to the users, vibration feedback, data displayed on dashboards, voice-interactions, playing context-aware music, etc.
- Smart sales. The end-users for Sales are the salesmen themselves. The results of analytics need to benefit the salesmen, and through them, sales. The desired results of analytics are recommendations related to what specific actions need to be taken by the sales organization and personnel in different situations. The analytics results need to be provided to salesmen on mobile devices, as is usually convenient to them.
- Smart healthcare. This domain has many users – patients, doctors, paramedical staff, administrative staff, etc. Each user has his/her own expectations from analytics, to enrich their experience in the end-to-end healthcare system. Patients expect to receive guidance and recommendations from the system more often than what doctors can give them. Doctors expect to receive pre-processed data, labelled as much as possible, to assist them in their work.
- Smart TVs. Who can overlook this – what do you call it? – device? gadget? Especially from its abundant potential for great UX. Many (most?) houses are TV-centric, though individual-lives are Smart phone centric. It is the interactions, the networking, the customizations, the personalization, etc. which is so very attractive about today’s Smart TVs. And intruder comes – it shows up on your Smart TV; about to rain – it shows up on your Smart TV. You can leave messages for others in your family on your Smart TV.
I have so far not even mentioned that grand finale of smart entity called Smart Cities. That is the culmination of many shades, extents, aspects of smartness in many products and services. More about that in perhaps another post. Watch this space.
Though I have only given many examples of three aspects – Data-availability, Analytics-capability, UX-design – things are much more complex. There are surely more aspects – like power efficiency, performance, cost, self-organizing, self-learning, etc. May be more about that in another post.
In general, though, the bigger the overlap of these aspects, the smarter your system is.
So, how smart is your smart system?
There is no answer to this question, at the moment. As in many things around us, it is the user’s perception which decides the answer. So for some more time, we need to live with the jargon – “smart”, “very smart”, “really smart”, “smarter” …