Robo-Collie, it’ll never work…or will it?

The news has recently been bleeting on about some scientific work done by Swansea University, suggesting that robots could replace sheepdogs to herd sheep. Of course this isn’t exactly what Swansea are suggesting, but just for a moment let us address this. The researchers have used GPS data to show how mathematically the sheepdog and the shepherd work together to herd a whole group of sheep successfully. Their conclusion is that there are two simple rules to it and that is that. But the main problem with this conclusion, as accurate as it may be, is that it is based on watching successful herding and not on watching how that successful herding was actually accomplished.

Having grown up around farms, I am quite aware of the tremendous skill and determination required to run a farm. Watching a sheepdog working, changing direction on a sixpence simply on a whistle command, is quite awe inspiring, especially when the effect of this on a sheep herd is instantaneous as well. This duet, or in some cases trio (with two dogs), work in perfect harmony to coerce a group of herd animals into a pen only just big enough to hold them. But the thing that comes across when watching this wonderful spectacle is not the mathematics of the movements, it is the skill of the animals and owner at reading the herd, in being agile and able to react on a moment to control what could otherwise result in a breakout.

Watching on from a mathematical point of view, it is no doubt possible to define the rules that were applied to make this well oiled machine work. But that is no more useful than defining how a stream may, over thousands of years, cut a course through  valley, only to be wrong because you didn’t realise there was a weakness in a line of rock half way down and this caused a completely different route. You see, by nature, sheep are unpredictable and so the dog and farmer control this through their experience and ability to judge and read the herd, rather than by just applying a rule and knowing it will work every time.

Now, the research is not actually denying this. What they have done is to use the rules observed from herding to suggest that robots could be used in other situations where here are large groups of people involved; crowd control or oil spills are the examples in the article. And this idea is actually very fascinating.

The Internet of Everything is the idea that any system will be ‘on the net’ and therefore measurable. The principle is that we not only collect the data from ‘everything’ but we also then interpret and react to it. This is an example of that. Scientists are looking at how one group, in this case a herd of sheep, react to herding and then applying this to another situation where a group may need to be herded in a similar way.

Of course the news reporters translate this into a small scale example, the person stuck in a dark room who can then be saved by a robot guide. That is small fry. So let’s extrapolate this into something rather more relevant.

Over the last couple of weeks the great British countryside has pretty much ground to a halt as millions of holiday-goers edge their way down the motorway to try and gain themselves their four inch plot of seaside heaven. The weight of traffic on one or two arterial roads is simply too much and so it takes eight hours to make a journey that on any other day might take half that. In the Internet of Everything world the cars will be herded, in real time, down the most suitable roads. They can be assigned a route based on availability, weight of traffic, capacity of the road, destination of other cars vs their car, the need to stop for food, water, a wee. All of these variables will be calculated by the vehicle, which will most likely to be doing all the driving anyway, and the roads will stay moving.

This is where the world is heading. Whether it be large groups all going to the same destination, or individuals trying to get somewhere whilst avoiding the crowd, it will be possible to use live data and mathematically rules to shepherd them in the most optimal routes. And that will be the power of big data as well. Take another salient example, the spread of the Ebola virus in Africa. Using data obtained using non-invasive medical devices, a patient will be able to be diagnosed in minutes, their presence will then be known ‘on the net’ and quarantine protocols updated accordingly. The ability to track the movement of the disease and then apply the rules we already know work will be much improved.

The power of data is impressive, and it will continue to impress us even more as we realise the potential for it’s use in the future. But an important part of using data is interpreting it in the right ways. The application of ‘intelligence’ to data, and not just generalising that one plus one must always equal two, will be how the Internet of Everything changes the world. Because the reality is that one plus one does not always equal two, sometimes it makes a window too!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s