Internet-of-things - everything disconnected?

Almost everyone has predicted the wave of Internet-of-things,  current prediction is 26 billion connected devices in 2020 (1). These devices are mainly sensors and actuators that are all communicating, sharing data and making everyone's life much better(?) - and now even Google have entered the market of sensors and actuators (ref the acquisition of Nest Labs (3))- really exciting if you ask me :-) And Google is on the right track!
For the consumer market, Internet-of-things has, so far, been mainly about how to implement smoke- and fire detectors, light- and heat control, window shades etc, all interconnected and ensuring that you can control and reduce the energy consumption in your house, provide better warning of any possible hazardous fires, even leakage of gas etc. It has also been about how to make your washing machine and refrigerator more "intelligent" by knowing what is needed to be picked up by your local grocer on your way home, even having an app that gives you a recipe for today's cooking based on what you have in your fridge!? All to best for the "common man" and to ensure that we become more cost efficient and secure, but has this really happened? Is everything really connected? No, surely not to a large extent and why?

One reason is of course cost of the devices, deployment/configuration and connectivity to the Internet. Surely the prevention of a fire or break-in in your house can justify these investments? Yes, but that is of course not necessarily what triggers the consumer, what triggers the consumers is if the infrastructure is installed and connected by a service provider, the local energy utility etc and that the consumer can be offered a subscription of the various services. This becomes even more evident when you look into sensors and actuators within healthcare. 
  
Globally, 87% of the over $2 trillion spent annually on healthcare goes to the cost of the care itself, such as the work doctors do, emergency room visits, medication, and procedures. The service provider can apply predictive analytics to the data acquired from the the sensors in each home and not only monitor elderly people's health condition in their homes, but actually pro-actively save life and ensure that they stay in their homes much longer to the benefit for the elderly themselves, for their family and for the society at large that will save millions of dollars in healthcare - talking about compelling business case!!
A wearable bracelet attached to the elderly patient is all what it takes, of course you can do more, there are hundreds of sensors in market (check out http://devices.wolfram.com/), but such a bracelet is a small investment with a major impact on the healthcare of elderly:

• Location awareness, the person can be tracked outside his/her home when the person actually should be inside his/her home (night) and the person can be tracked being on the very same spot for a long time (laying helpless on the floor etc).
• Measuring blood sugar as well as pulse and pressure – pro-active prevention of  heart attack by heart rate pattern recognition (predictive analytics).
• Measure too high temperature from the patient as well as the environment (fire), also measuring humidity due to some accident (water leakages etc) in the home and/or in the bed (urine).

Is this a business opportunity - is it a global trend? Yes, you bet:

IN 2016 The U.S. market for

wearable smart fitness devices

will exceed $1 billion (2)

So lets ensure that these simple sensors are being implemented to the benefit of the elderly and society at large - lets make the Internet-of-things become really connected!

Reference/source:
1) http://www.computerworld.com/s/article/9245507/Why_the_Internet_of_Things_may_never_happen
2) Parks Associates 
3) Google & Nest labs

 

Industrial Internet and cognitive computing

Internet has been an industry revolution; the second after Ford revolutionized the car production more than a hundred years ago. Internet has also been a major disruption to multiple industries, not just the telecommunication industry, but indeed through the way we work every day across all industries, the way we share data, knowledge, skills and experiences, thus it has been disruptive across multiple industries - its going to take a few years before we experience anything like the impact of Internet again.
Currently we are now experiencing a paradigm shift in the way that we are leveraging the Internet technology, its even a disruption to some technologies and services. Just think about the various concepts and technologies that the IT industry and their media all are talking about; Internet-of-things, Big data & Analytics, Cognitive Computing, Smartphones, predictive analytics - much of it has been around a while, but its all about the same thing, its all about processing of humongous amount data, unstructured and uncertain data acquired from social media and "sensors" that can feel, see, sense, react and learn, its all about how the Internet is being industrialized! So, is this great, is this beneficial to the global business? Yes it is because it represents potential for more revenue for less cost.

What is the driver behind this, is it the data? No, not data alone, its the capability of applying logic to the data analysis, read capability of applying knowledge, skills and experiences to the data analysis, its about Big data and Analytics! I mentioned earlier that this is a paradigm shift that is "..disruption to some technologies and services", interestingly is that Analytics is now disruptive even to the consultancy services as consultants now distill insight out of big data and brings analytics separate from traditional project based models and creates more business value to their clients than ever before and to a significant lower cost through reduction of risk through learning - its like in the human world, applied knowledge, skills and experience reduces risk - if you have already experienced the case or the project task ahead, you are most likely able to reduce and mitigate the risk as much as possible to make sure the task or project is being successfully implemented.

So why is this a paradigm shift? Its a paradigm shift because up until recent years we have only applied analytics to structured data, we have been missing appropriate algorithms to deal with uncertain- and unstructured data, moreover we have been lacking cognitive capabilities to ensure we make the computers take advantage of the information being fed into the system, we have been lacking the capability of making the machines learn (cognitive computing) from their success and mistakes!
Why is that important? Its important because the analytics we apply to the data will provide us better answers/results and we will be able to ingest these answers or results back into the system as part of the cognitive process and the system will learn from its own experiences. Recently IBM has shown us how this can be solved through their Watson technology that was used to beat the Jeopardy masters.

This was a major achievement because it demonstrates how a computer-system can leverage massive amount of unstructured data and learn from its processing achievement, thus without manual intervention the system will, in real time, improve its hit rate the more questions being asked. How can we apply this in the real world? What humongous amount of unstructured data can we feed into a cognitive system like Watson that will give significant business revenue to a lower cost? Back to the Industrial Internet and Internet-of-things. Sensors have been implemented on large engineering machines like mining machines, large ship engines, locomotives, airplanes, power grid, transformers etc. Machines that are highly complex and to a large degree "revenue-sensitive" as any disruptions will cause significant loss of revenue and/or contract compensation claims. The purpose of these multifunction sensors is to ensure that maintenance can be more pro-active, more predictive through monitoring these data and applying the operators skills and experiences to ensure actions can be taken pro-active. 
But what if all this gets automated, what if we ingest all the operator skills and experiences through manuals and disruption/root cause reports etc into a system that applies all this knowledge, skills and experiences to the uncertain sensor data from the entire cluster of the mining machines globally or all the power plants globally? 

Such a system can be the ultimate oracle and subject-matter-expert (SME) to be consulted when incidents happens. We could thus be able to both give the operators the correct answers to incidents as well as learn and predict where maintenance is needed pro-actively before the machinery disrupt the production. This is a multimillion dollar business where there are significant opportunities to make the engineering industry more competitive and cost effective - and the big players are currently exploring this opportunity.
 
On top of this, we can reduce human suffering and even life as disruptions to some of these large machines for sure also threatens human life. Talking of which, lets see how the Industrial Internet applies to the health industry. We all know that we will find massive amount sensors in the surgeons operating theater, sensors that are monitoring the patient being treated - be sure we can improve the monitoring significantly by applying analytics to help the surgeon to take the right decisions during the surgery, but we can also look into simpler use cases like helping the clinicians in the review and authorization of medical procedures and treatment simply by ingesting patient medical records into a cognitive system like Watson and perform the analysis to identify the missing info to set the correct diagnosis.

20% of all cancer cases receive the wrong diagnosis initially - in some cases it is as high as 40%. I personally have experienced a wrong cancer diagnosis, luckily I survived just by coincidence. Health care systems struggle daily with rising costs and uneven quality, despite hard work of well-intentioned clinicians. Someone has calculated that only 1% improvements in healthcare yields $60B savings globally!
And if this is not enough, IBM is now about to set up Watson in their Research lab in Kenya for the purpose of providing life improvement cancer Research;

"Women in sub-Saharan Africa account for 22 percent of all cases of cervical cancer worldwide mainly due to a lack of services and knowledge. Watson could provide new insights into the evolution of cervical cancer in Africa and suggest new approaches for its prevention, diagnosis and treatment. By feeding back valuable clinical data about their field observations, healthcare works will be able to contribute to improving Watson’s inference abilities"

- and within education;

"Currently, half of African children will reach their adolescent years unable to read, write or perform basic numeric tasks. The key to improving these statistics is a thorough understanding of student performance, teacher expertise, attendance levels, class sizes, linguistic abilities and learning materials. While previous information systems have only provided a limited view of point problems, using Watson technologies, Center of Excellence for Data-Driven Development (CEDD) aims to create new holistic approaches for analyzing data to identify previously unrecorded correlations. For example, Watson could identify the link between a contaminated water borehole, an epidemic of cholera and the subsequent low levels of school attendance in the region. Watson could also help to uncover other causes of low school attendance in a particular region such as a lack of sanitary supplies and cultural traditions placing childcare responsibility on older siblings."

References/sources:
Marco Annunziata: Welcome to the age of the industrial internet
GE’s Industrial Internet and the Smart Grid in the Cloud 
IBM's Watson Gets Its First Piece Of Business In Healthcare 
IBM’s Watson starts its medical career 
IBMs Watson group announcement 
Watson goes to Africa