Books

Books
Books written by Ray Sullivan
Showing posts with label IBM. Show all posts
Showing posts with label IBM. Show all posts

Saturday, 4 May 2013

Leaving Virtual Breadcrumbs

IBM have carried out a major study of mobile phone traffic in Abidjan, the largest city on the Ivory Coast.  They analysed over 2.5 billion mobile calls from the five million mobile phone users on the Ivory Coast as part of the study.  They didn't have access to consistent GPS information but were able to identify the closest phone mast to each call as it was made and from that they drew a population map of the city that allowed them to make a prediction about traffic routes that the city planners hadn't identified.

It sounds crazy to think that mobile phones can tell us more than census returns or even good old fashioned observation of where the houses are built but the obvious weakness in those metrics is that it's people, not houses, that travel on buses.  And people are where they need to be, when they need to be.  Which may be at work, at play or at home.  And generally with their mobile phones.

What IBM found was that the city needed two more bus routes laying on and that the net result of that would be a reduction of travel times of ten percent.  Or put it another way, a measurable efficiency saving.

It's not the creation of two new bus routes that most of us will never see, let alone enjoy, that's really significant here, though, it's the methods developed to analyse such large amounts of data to make that analysis.  It's a mobile world, and not just in the first world.  I don't know a great deal about the Ivory Coast, but a cursory glance at the main economic and demographic data would suggest it is a relatively stable region of western Africa and that the adult population numbers about ten million.  Or to put it another way, there's about one mobile phone per household, similar to the situation in the UK about 15, maybe 20 years ago.  The IBM report also suggests that public transport is also still very much prevalent.

Now I guess one of the questions that should be asked is whether the commuters who used the bus system knew that they needed an additional two bus routes, and if they did, had they asked  for them.  Because that's probably a simpler, low tech solution to this kind of problem.  It reminds me of my early days in the RAF where I learned that new bases were established without formal paths between buildings - instead they grassed all the potential pathways and waited to see where the natural paths were formed.  Any two airmen may have had opposing views on where the paths should go, but stuff a thousand random guys and gals on a base and avoid telling them where to walk and they'll work out the most efficient routes without needing a scale map.

But underneath this is the prospect that we are all leaving bread crumbs behind us wherever we go.  If it's public transport we need to manage, look at the crumbs.  If we have half empty buses but full routes into town, maybe our prices or timings are wrong - as a group, we consumers will gravitate to the most appropriate deal going and might, just, relinquish the convenience of our cars if the benefits outweigh the negatives.

If a lot of people are hanging around a bus stop consistently between certain times it suggests that they could change the timings, too.  How about capturing those people who are hanging around town - maybe waiting for those pesky buses?  Perhaps local bars and cafes could use that information to change their opening hours or to make intelligent offers to draw punters in.

Maybe towns and cities could look at  where people wander, to decide where they need to put information, to direct them to where they will spend money, maybe to direct them from less salubrious parts of town to protect them?

The point is, wherever you go with your mobile phone, you leave breadcrumbs.  When we all access the same virtual loaf, we leave real data.  That data is worth more than a loaf of bread, it's a bakery and the product from its ovens is knowledge.


------------------------------------------------------------------------------------


                                                          Visit my Book Website here

        Visit Project: Evil Website here                                        Visit DLF Website here

        Follow me on Twitter  - @RayASullivan

        Join me on Facebook -  use raysullivan.novels@yahoo.com to find me

Thursday, 25 April 2013

Samsung Has Its Thinking Cap On

One of the side events I witnessed at the Gadget Show Live recently was a demonstration of a thought controlled roller skateboard, demonstrated by one of the founding fathers of Twitter.  Putting aside any inclinations towards scepticism - the thought control element could have been kidology, it may have been controlled by a remote control out of sight of the thousand or so spectators - it was shown to drive around the small arena, accelerating and decelerating as needed to negotiate the circuit.

It worked reasonably well - it was driven onto the main stage, down a ramp and around the front seats on three sides of the stage, with the driver crouching down and slapping kids on the palm as he swept past.  The concept is that you train the computer - a Windows 8 tablet on this occasion, presumably because the Gadget Show Live was sponsored by Windows 8 although it is understood that a Samsung Android tablet has been used in earlier iterations - using a headset that is sensitive to neuron activity in the brain.

It appears that each driver has to train the computer to recognise the way their brain generates neuron activity. We're all different, apparently, so consequently a pre-programmed device ready for fine tuning is out of the question.  This sounds remarkably similar to the issues and challenges overcome by the teams training rats to push buttons using brainwaves, as mentioned on my blog posting from the 10 March this year 'Leave the Mouse, Get a Rat.'  So, for brainwave controlled activities we're about as effective as rats.  That's reassuring.

Samsung have been working with MIT to develop this technology further  using a headset bristling with EEG measuring electrodes.  Think of a swimming cap with a poor man's dreadlocks and you get the idea.  Hooking up to one of their Galaxy tablets they've had a fair bit of success in selecting and launching apps.  As one commentator states, thought controlled technology will be a boon for those with mobility issues, and persons suffering with illnesses such as Locked In Syndrome may have some relief.  Looking beyond low hanging fruit such as that, it also presents a wealth of opportunities for those of us lucky enough to not be classed as disabled.

Controlling the TV and the DVR by thought control has to be an aim, although the resulting carnage in houses up and down the country needs to be considered as TV channels are changed in the literal blink of an eye.  Adding an extra input dimension to operating your computer has to be an objective, too.  As we demand more from our programs, the need to manipulate needs more than a virtual extra pair of hands.

Part of the tests carried out at MIT is using the thought control to manipulate a music player, getting the human equivalent of the lab rat to select, play and pause classical music tracks.  At present the accuracy of such tests is between 80% and 95%, which isn't perfect - I would expect around 98% accuracy using conventional controls by persons familiar with the software.  However it is probably a lot better than most would expect.  The researchers are very happy with the results and are looking at ways to make the headsets more convenient, such as replacing the current wet electrode requirement with a dry electrode.

It's early days, but if a viable range of controls are developed then maybe the sensors will be fitted subcutaneously, allowing computers and other devices to be controlled just by thought.  Like Google Glass, this technology has the potential to change the game permanently and my guess is the days of clunky rubber headgear are limited.  For this application, anyway. The technology will undoubtedly develop faster as the results improve, and I expect the progress to increase in leaps and bounds as the capability is realised.  And of course it's not just Samsung looking onto this technology, IBM have a research project working on it, so we're looking at some heavyweight research going on.

In my opinion, if any technology is worth thinking about, this is it.

------------------------------------------------------------------------------------


                                                          Visit my Book Website here

        Visit Project: Evil Website here                                        Visit DLF Website here

        Follow me on Twitter  - @RayASullivan

        Join me on Facebook -  use raysullivan.novels@yahoo.com to find me