Paul Caldwell Presents the Internet of Things and Big Data

At the April Monthly Meeting, Paul Caldwell presented a far-reaching overview of the latest consumer and professional IT technology. His excellent and detailed PowerPoint presentation was entitled The Internet of “Things” and you can download and view the PDF of it here.

Paul started by recounting Moore’s Law, first stated in 1965 – that computing resources double every two years. To grasp the exponential effect of constant doubling, we might recall the story of the Chinese emperor hoping to reward his best general by placing one grain of rice on the first square of a checkerboard, two on the second, four on the third and so on (slide 18). The 64th square would have contained more rice than there was in China!

Applying this model to the communication/IT scene (slide 3), we could say that it started in the 19th century with the telegraph and eventually evolved into the extensive scene of today, with such an incredible amount of data and computing resources that, in recent times, no one person can hope be a complete expert on anything.

Arguably, that tipping point was reached in 2007 when Steve Jobs brought us the iPhone. So much more than a mobile phone, it is loaded with many types of sensor (slide 6) – including those for location, acceleration, temperature, and light – together with matching application programs – Apps – also including those for Texting, Facebook, Skype, YouTube, news-feeds and games – which present friendly user interfaces for all those resources.

Another recent arrival is the Personal Heath Monitor such as Fitbit which can monitor your daily, steps, calories burnt, heartbeat, pulse, temperature and other parameters. That data can be processed in many ways and can extend to detecting signs of a heart attack up to a month in advance.

Today’s Smart Home (slide 13, of which Paul’s is one) can have intelligent and remotely controllable air conditioning, lighting, security monitoring, refrigeration, vacuum cleaning, cooking, coffee brewing and more. It does this internally via WiFi, Bluetooth and RFID, while external commands typically arrive via 4G from a smartphone. Paul set up his own Smart Home system himself (slide 29) but firms like Stuff Fibre, who work with Noel Leeming, can design and install a complete, integrated system for you.

The use of smart technology also extends into Industry. Paul cited the case of his father who was an engineer at NZ Steel. He was an expert in the “smart technology” of the day, which included the use of programmable controllers, though that programming mostly had to be done on a device-by-device basis, directly involving switches and microswitches. Nowadays, everything is digital and highly programmable in very sophisticated ways, able to flexibly respond to the full range of information that the steel mill generates.

It’s a broadly similar story with supermarkets and large stores. They no longer do periodic stock takes. Rather, they now use supply-chain management, where all products and all shelves are continuously monitored and automatically reordered and restocked.

On the road, there will soon be parking sensors that can tell you where there is an available parking space and, of course, in the typical modern car, there are many sensors whose data the owner, the mechanic and the dealer alike can benefit from.

But the promise of Big Data (slide 16) extends far beyond even this. Most of us will recall that IBM’s Watson, using Artificial Intelligence was also able to access a full range of databases in real¬† time and beat the best human contestants in the quiz game of Jeopardy. Watson is now being used in much more important fields, including Cancer Research. In this field alone, 5000 papers are published every day! This is far beyond human ability to follow but Watson is able to scan every piece of data and match it with human specialist case notes. So Watson might “see” that there is a trial going on in Budapest, where people with the same symptoms as a particular current patient responded well to such-and-such a treatment/drug and immediately advise his/her doctor about this.

On the wider scale of epidemic control, Big Data has also played a part. As we know, Google looks at what you are searching for. With influenza epidemics,¬† doctors used to fill-in forms but that typically resulted in a month’s delay. By looking at the searches people were making at any given location – typically for flu treatments – Google was able to detect the real-time spread of the epidemic. That worked until everyone became aware of this method and mimicked it, thereby confusing the algorithm!

Paul gave a number of demos of the “toys” he brought with him, including a WiFi enabled switch (which might turn on a light) that can be activated in response to a WiFi signal triggered in turn from a remote smartphone.

However, Paul’s biggest plaything was surely “Alexa” – outwardly a small, pie-sized object sitting on the table (slides 27 & 28). It – or rather she – could recognise Paul’s voice and, as we soon learned, those of others in the audience as well. Although Alexa would only obey Paul’s pre-set commands (such as to send an alert to someone or to add an item to the Caldwells’ shopping list), she would attempt to answer any questions including from members. On Paul’s request, she even told a joke or two, which brought a few smiles but then a member asked her whether she liked Paul. Her reply – “I don’t have an opinion on that” – produced the biggest laugh of the meeting. Alexa is not the only such device – Apple’s Siri, Microsoft’s Cortana and Google’s Voice are among its competitors.

Needless to say, Paul’s outdoor “toys” include a drone which he has used to photograph his beach house from the air. Less of a toy is WAZE (slide 23) – the world’s largest community-based traffic and navigation app which Paul routinely uses in the rush hour, instead of listening to the official – and usually out of date – traffic reports.

Paul also explained how Big Data has substantially replaced phone surveys. The latter are inherently biassed because they include only those with landlines and who have almost endless patience. So pollsters now prefer mining on-line data, including what social media and news sites are publishing, and are able to draw surprisingly accurate predictions of how people are likely to behave. Those very much include political trends and voting patterns worldwide but, notably, the recent US elections. Paul was one of the few observers who was not surprised by Mr Trump’s victory and that, he contended, was because the Trump campaign made full use of the above techniques. By contrast, even though Mrs Clinton had the support of most of Silicon Valley, her people did not apparently make good use of the same methods. Those methods can be enhanced by the use of FUD – Fear Uncertainty and Doubt: stories beat facts (slide 21).

Nonetheless, to make constructive use of Big Data, the data pyramid (slide 20) must be made to work. At its base is the raw data itself. Next is Information, then Knowledge, followed by Understanding. At the apex is Wisdom or Evaluated Understanding as we might describe it. But again, the obvious challenge is to make that pyramid actually function as the slide depicts.

On a more mundane level, Paul noted that each of us could escape the email traps that the ISP’s effectively set when they tie you to their own email servers, with addresses like j.smith@xtra.co.nz or j.smith@vodafone.co.nz and which use POP or IMAP addresses. However, you may get your own family domain name (e.g. smith.co.nz) for about $70 per year. Whether you use it to create your own website or not, you can certainly use it to generate typically up to 10 email addresses (e.g. john@smith.co.nz) for yourself and family members. Those emails would then be on an exchange email server, which is much more secure and flexible than POP or IMAP based ones. If you don’t want to go as far as creating your own domain, you might prefer to use Outlook.com, which is free and exchange server based.

At the end this long but engrossing presentation, Paul received a well-deserved ovation and stayed on talking to individual members. Thanks, Paul!

Wayne Power


Leave a Reply