Bench Talk

The Internet of Things (IoT) will provide an avalanche of data. Remote sensors for just about every ‘Thing’ will provide data on the location of cars, busses, packages, and your dog, as well as measurements of your heart rate, blood oxygen level, and even the amount of half and half in your refrigerator. Lots and lots of data.

Data on its own is not really that valuable, however. What we need the IoT to do is to help us organize, analyze and act on data in ways that provide, in order of increasing value, information, intelligence and wisdom. Let me explain each of these in more detail and show how they build on each other creating different value 'layers'.

Data is the simplest and familiar concept. Data can be considered a single metric, perhaps a GPS location, temperature or light level. Data is useful and even single data points are helpful when you need to adjust the back lighting level in a smart phone, find your house keys or adjust the fan speed in a communications chassis.

Information results from organizing data, over some 'axis', such as time or location. For example, many data elements from an electrocardiogram (EKG) monitor, taken over time, can be organized to determine a heart rate. Measurements from ambient light sensors over an outdoor parking lot can be used to determine if lighting levels are adequate, or if lights need to be adjusted or serviced. You might say that information is the 'integral' of data over an 'axis', often measured via time or location. Even in these simple examples the information is clearly more valuable than the individual data measurements.

Intelligence results from operating on information, usually by applying a process or algorithm. As an example, the EKG measurements mentioned previously were easily organized to determine a heart rate. If an algorithm is applied to the Information available from the EKG, even more interesting and valuable Intelligence can result. An Automatic Electronic Defibrillator, or AED, can use EKG information to determine if a heart is undergoing a cardiac event. It can then determine the best time and waveform to use to 'shock' the heart back into normal operation.  Clearly the application of algorithms to EKG information creates significant additional value. We could say Intelligence results when we apply a useful ‘function’ to Information.

Creating Wisdom in the IoT requires data, lots and lots of data (IStock)

One might think that Intelligence is as far as we can take data up the value chain. It seems possible however, in the best of all possible IoT worlds, that an additional level of value can be developed above Intelligence. Let's take the EKG example one step further by assuming a large set of Intelligence is available. If Information from every AED and EKG was available (I know, very unrealistic today, but as I said, let's assume the best of all possible IoT worlds) the Intelligence algorithm could be 'tuned' to perform in a wide variety of situations, and used for something difficult to predict when the original algorithm was developed. For example, perhaps a tuned algorithm could analyze EKG information to detect a secondary and faint heartbeat, indicating that the patient is pregnant. This improved algorithm could then adjust treatment accordingly. EKG measurements from a variety of pregnant patients would need to be available to tune the algorithm successfully.

In another example of increased value, let's assume that a medical monitor has been routinely worn by someone who is having a cardiac event and is being treated with an AED The medical monitor could provide critical Intelligence to the AED that could be used to modify AED operation. The nature of the cardiac event could be quickly categorized and the type of 'shock' applied optimized. If pre-event intelligence was available for every patient, algorithms could be further refined. Even more importantly, perhaps preventative treatment could be prescribed if the medical monitor could identify early warning signs of a probable cardiac event.

These types of medical monitors are becoming available and could eventually be used in similar scenarios as described above. A recent announcement from ST Microelectronics shows the types of capabilities their STM32F MCUs can deliver in advanced medical monitoring applications. Algorithms become increasingly valuable in the IoT ecosystem and can add significant value to the underlying hardware.

Algorithm improvement, tuning and extension into things like preventative prescriptions creates another level of value, one we could perhaps call Wisdom. Using an extensive history algorithm can transform experience into predictive action, by knowing ahead of time what will happen, to prevent or mitigate an event that requires deep knowledge and experience. This level of experience, in the IoT, only comes when intelligence can be combined from many, many sources, and advanced learning algorithms are employed. Wisdom in the IoT can perhaps be thought of as applying deep learning on the historical performance of algorithms over very large sets of results.

Widely available 'Big Data', cloud storage, ubiquitous connections and many other elements will need to fit together relatively seamlessly to realize the best of all possible IoT worlds. This world however, promises to be one where Wisdom can be discovered and deployed to dramatically improve lives in completely unexpected ways.