Monthly Archives

10 Articles

Posted by on

Project Santander: Sticking Data in the Database

WARNING: TECHIE POST. Following my visit to the smart city project in Santander with Telefonica last year, I was inspired to start building my own smart home based on similar technologies. This is partly an exercise of my rusty engineering skills but mostly about learning the realities of smart cities/smart homes through experience. I’ll write up the lessons in a much less techie form, but for those who are interested, I’ll also be documenting the detail here.

In the last episode I got my first sensor up and running on the end of an Ethernet cable thanks to the RESTDuino sketch. Now I need to get the data this send back into my SQL database.

This proved to be incredibly simple. Now note: I am NOT a coder and this stuff is probably seriously ugly. But it works and that’s what I care about right now.

Into the code I had put together for the AlertMe API I added the Pest library, which makes accessing a RESTful interface using PHP dead easy. It’s then a simple task to extract the useful data from the little JSON string that comes back from my light sensor and squirt it into a new table in my SQL database.

It’s worth talking a little bit about the database here. I’m structuring it with a series of tables for each different type of data: power, temperature, light, etc etc. Rather than one for each room. This should mean that I don’t have to create a new table each time I add a new sensor. I can just record values indexed by their time stamp and room code. I’m hoping this should make things simpler.

Anyway, after just a few failed attempts (mostly down to my lack of understanding of JSON and arrays) my code was working and happily sticking data from the light sensor into my database.

Next!

Posted by on

Project Santander: Ethernet and REST

WARNING: TECHIE POST. Following my visit to the smart city project in Santander with Telefonica last year, I was inspired to start building my own smart home based on similar technologies. This is partly an exercise of my rusty engineering skills but mostly about learning the realities of smart cities/smart homes through experience. I’ll write up the lessons in a much less techie form, but for those who are interested, I’ll also be documenting the detail here.

Excitement today. My Ethernet shield finally arrived from China. What can I say? I’m a cheapskate and I don’t mind waiting a little longer to save a few quid.

So, straight into testing this evening. I’ve decided to start with the RESTDuino example as the basis for my sketch. So I downloaded the relevant files and uploaded them to my test Arduino with the Ethernet shield attached. Note: first challenge, the sensor shield won’t fit on top of the Ethernet shield nor will they fit the other way around because the ICSP header gets in the way. Looks like I might be going down the prototype shield route to attach all my various sensors.

Anyway, I left the defaults as they were in the demo sketch and uploaded it to my board. I plugged an ethernet cable into the nearest wall port (I have more thanks to the fantastic Devolo dLAN kit I am currently testing) and hooked up an LED to pin 9 as suggested in the demo. I then hit the address — 192.168.1.155/9/HIGH — that should have turned it on.

Nada.

So a quick scan of the network using iNet, one of the most useful iPhone apps I own. Sure enough this showed me that the Arduino was on 192.168.1.38. Despite it being assigned a fixed IP, it is clearly getting one on dynamic assignment. There’s something to sort down the line.

Anyway, stick the new address in — 192.168.1.38/9/HIGH and sure enough, the LED lights up. Hoorah!

So, let’s get a bit more sophisticated. A sensor reading.

Back to my favourite sensor (OK the only one I had handy at the start of this exercise), the photocell. I hooked this up in a voltage divider arrangement with a 10K resistor and fed the signal line back to analogue pin 1.

Go to 192.168.1.38/a1 and what do I get? A JSON string that looks like this:

{“A1“:”737“}

Superb! Even I ought to be able to work out how to pass than into my SQL database. But that’s for next time…

Posted by on

ILM: The Future of Management

For the past few months I have been blogging for the Institute of Leadership and Management on ‘the future of management’. In these blogs I have tackled a wide range of subjects, from the infamous pivot of start-up culture, to the need for skills over knowledge. I thought it about time I collected these various posts in a single index on this site, including the most recent one that went live this week on ‘Five Skills for Tomorrow’s Manager’. So here they are in (roughly) reverse chronological order. Look out for the next post next week.

Five Skills for Tomorrow’s Manager

Four Ways to Improve Innovation

Managing the Bionic Worker

Does It Pay to be Nice?

Manage It? Measure It.

IT Managers: Leaders of the Future?

The Naked Manager

It’s Not What You Know, It’s Knowing Everyone

Get Out of Your Inbox: It’s Good to Talk (and Show, and Share)

Why Time isn’t Money: Getting Over Presenteeism

NSFW: The Oversharing Generation at Work

Future of Management: Why Facts Will Be Worthless to Managers

Pivot: Why Tomorrow’s Leaders Must Be Light on their Feet

The Future of Management: Managing Without Employees

The Future of Management: Computer Says Do

Posted by on

Lovejoy: “Will there be enough jobs for everyone?”

Workaholics of the world, be warned. In the future you might need to find a hobby (or two).

Twice I’ve been on Sunday Brunch. Twice Tim Lovejoy has asked me the same question: “Will there be enough jobs for everyone in the future?”

My answers have been fairly flippant: I don’t think there would be enough work to fill five days for most people. The more I think about this though, the more I think it has to be true. I haven’t had the resource to look at it in depth yet, but I was inspired to have a first stab at tackling the question properly by the current debate around raising the minimum wage.

So just for you Tim, here’s my current thinking (to be refined).

Today

The first thing to point out when trying to answer this question is that it isn’t a future problem: there aren’t enough jobs today. According to the latest figures from the Office for National Statistics there are 2.39 million, unemployed but economically active (i.e. available to work) people over 16 in the UK. Contrast this with the number of available vacancies — 562,000 — and you see that already we have nowhere near enough jobs available for everyone to be able to work.

The situation is actually more complex than this, as you might expect: a further 1.46 million people who are technically employed are only in part-time jobs when really they want to work full time. It’s hard to estimate how many extra jobs/job hours would be needed to create a situation of full employment. But if you take each part-time job as being 50% of the hours of full time (this feel generous but I haven’t found any good data to work from) then we would need more than 3 million new jobs to fully employ everyone.

We are slowly crawling our way out of recession of course, which doesn’t help the vacancy numbers. But it’s important to point out that the unemployment rate has rarely dipped under 5% in the last 30 years and for much of that time it has been significantly higher. You need a certain number of people looking for work so that you can fill vacancies and companies can grow (though clearly you don’t want anyone to be unemployed in the long term).

Pressure Points

When looking to the future we tend to start by looking at the pressure points in a market or around a particular issue. There are plenty of pressure points affecting unemployment.

First of all, we are simply living longer, healthier lives. More of us remain available to work, later in our lives than ever before. And many of us, certainly in senior positions, may have no problem carrying on: if the work is mentally stimulating rather than physically challenging, then why not continue working?

The state pension age has already been bumped once, it is unlikely to remain where it is with the pressures the ageing population places on government spending on health and welfare. Remember that over half of the welfare bill goes on pensioners already. That’s before you take into account the costs of the NHS and nursing homes.

As a nation we need to keep people active — both physically and economically — for longer.

Our lengthening lives are the primary contributor to a growing global population: there will be 9bn of us by the end of the century based on the median estimates from the UN. This increases the demand for products and services, but the number of workers required to deliver each product or service is declining. On a global scale, whether or not there will be enough jobs between now and peak population is in some ways just a question of which curve is steeper.

Looking more locally you have to deal with the thorny issue of immigration. Within the EU there’s good evidence that migration tends to be temporary — more than half of the number who arrived between 2004 and 2009 went home within the same period, with more leaving than arriving in 2009 during the downturn (figures from the ONS). Understandably people come if there are jobs and go away again if the jobs go away. There’s no doubting this creates competition for positions, but bear in mind the people leaving the UK to find jobs or just live elsewhere. Net migration is under 200,000 people per year and the government wants to get it down under 100,000.

Finally there’s education. Much as I loved my time at university, I don’t believe the traditional 3–4 year degree has much of a future. The fast pace of the jobs and skills market means that for all the efforts of the universities, a degree no longer confers greatly increased employability. Many employers still use degree status and score as a handy filter for whittling down the number of applications to a manageable shortlist, but more and more employers I talk to are taking apprentices on at 18 or even 16 and shaping their skills themselves rather than leaving it to a university. Once you take into account the increasing cost of a degree, and the reduced chance of high earnings, going to university will lose its appeal for many school leavers who might instead choose to study at points throughout their career as required.

That’s potentially hundreds of thousands more 18–21 year olds who might be on the job market in a few years’ time.

Technology Trends

In short then, it seems likely there will be more job seekers in the future. But what about the number of jobs? This seems to be the crux of Tim’s question.

The history of human technology is a history of our ability to build machines that do things better than we can. There remain many things today that people can do better or cheaper, or that remain uniquely human skills. But this list is shrinking. Automation has come to almost every industry: factories use robots for repetitive tasks, customer service happens online and via interactive telephone systems, banks have replaced staff with kiosks.

In the developed world, people are expensive and risky when weighed against machines.

In the developing world, people are sadly less expensive and risks are more accepted: around the world humans work in appalling conditions for low wages to collect raw materials and produce goods. For a while, this lower cost and higher risk threshold will keep people in jobs. But it can only last so long: ultimately there will be cheaper and more efficient — as well as safer — ways to do many jobs. As livings standards improve and wages rise, more and more jobs will be automated.

The jobs will also move — or at least the machines will. The rise of on-demand manufacturing could seriously change the economics of many consumer goods, making it cheaper to produce a t-shirt to exact size and design specifications from raw materials, on the high street in Manchester, rather than ship many variations of the finished item half-way across the world.

This trend will be supported by the increasing cost of raw materials and energy: why risk wasting some of your decreasing margin on over-production when instead you can have the money in the bank before you make the goods? Any surplus is just a raw material that can easily be traded at market price.

Automation isn’t limited to the manufacturing industry. Knowledge jobs can and are being automated. Just look at the change in accounting software over the last ten years: as a small business or self-employed person it has become very easy to run your own accounts without recourse to an accountant. I still use one because I value their advice, because I hate filling out forms and because the prospect of screwing up with HMRC terrifies me, but many companies I know do not. Legal, HR and other professional jobs will be diminished if not replaced completely with automation. What’s the hottest topic in marketing right now? Automation. It is coming to every sector.

The Three Day Week

So in short Tim, no there won’t be enough jobs to go around. The issue is one of supply, demand and simple economics. The increasing population of economically active people will increase the demand for products and services. But it will also increase the number of available workers at a time when the number of workers required to deliver each product or service is falling.

Demand for workers is falling because the supply side of work is being supplemented by increasing automation — essentially robots, be they software or hardware. Automation will happen faster than demand increases due to population growth, and by the end of the century population will start to fall, expanding the gap between the number of jobs and the number of workers.

Which leads to a difficult question: what are we all going to do? But that’s a post for another day.

Posted by on

The Value of Ideas

Ideas are mostly worthless. This is something I find myself thinking every time I talk to a wannabe entrepreneur. Ideas gain value when they are applied. Entrepreneurs aren’t people with great ideas, they are people who apply the ideas they have. Or as Edison supposedly put it, so much more eloquently: genius is one percent inspiration, ninety nine percent perspiration.

All that said, the application of an idea in this internet age can be apparently almost trivial. I’ve spent the last couple of hours with an entrepreneur who has a very simple idea, albeit one that he has had trouble articulating. Talking to him a couple of weeks ago I felt there was probably something in what he was saying. Two hours of thrashing the idea around today and I’m now convinced there’s something in it.

What he wants to do is relatively simple: take large volumes of publicly available data and manipulate them into a useful, appealing format. It’s an idea of its time, very much in keeping with the current ‘big data’ trend.

The ingredients of his innovation are neither new nor unique: they are available to anyone in the world with some basic development skills. It’s taken him eighteen months (without any development skills of his own) to get to this point and there’s no doubting he has worked hard. But the product itself when fully live is relatively simple.

And yet no-one has done what he’s doing.

Ideas unapplied are worthless. But the state of technology today means that some of the best, simplest, original ideas can be applied quickly and with limited effort. It still takes a lot of perspiration to make them a real success, but once the effort of application begins, the value of ideas can scale very quickly.

Posted by on

BBC News Channel: CES-Why Wearable Tech Will Be A Success

I’m pretty confident in the future of wearable technology — at least the right wearable technology. I’ll be expanding on this on the Beeb later today, but here’s my recipe for what will be successful.

Functional

We all know that technological capability is no recipe for guaranteed success. But it’s certainly part of the equation. The first thing wearable technologies have to do is find strong applications.

Let’s be clear about what we mean by ‘strong’ though: the benefit has to be proportional to the additional cost over and above non-smart alternatives, whether the additional smartness is built into tech or clothing/accessories, and assuming all other factors are equal. Adding vast numbers of features to try to skew this equation (I’m looking at you Samsung Galaxy Gear) will not work. Adding too many features actually makes the device LESS functional by complicating the user interface and bulking out the design.

I think LG has the right idea with its Heart Rate Earphones that add useful fitness tracking functions to a device you might be using while exercising anyway. As long as the cost is within a sensible margin of headphones without that tech, and they function well as headphones (and of course look good), then I think devices like this will be appealing.

What I believe will be successful will be simple, one or two function devices, that look attractive (below) and have a very simple, natural, almost invisible user interface.

Desirable

Technology is increasingly invisible. Between its small size and more human user interface design, it is less and less obvious when something is a piece of technology as we used to think of it. Go back forty years and computers were hulking machines that needed you to speak their language. Now we stroke tiny devices and they respond to our touch, command and even hand gestures. It’s not a massive leap to see technology just disappearing as a discrete thing altogether and becoming part of the fabric of the world around us.
When it does so, the function will be increasingly expected and the differentiation will be design. Sound like any industry you know? Fashion.

Brand and aesthetics are incredibly important in wearable tech. As I talked about in my roundup from the closest thing Europe has to CES, IFA, last year, some companies really get this. If they’re not partnering up with well recognised style brands, they are employing designers from that world and working hard to give their brands greater appeal.

Normal

Ultimately the ubiquitous success of wearables will come when they become normal. When doing the things that only wearables can do becomes expected.

Health monitoring is the first good example of an application that might become expected. It may seem a little bit big brother, but collecting and sharing data about our wellbeing might be in all our interests, as taxpayers and as individuals.

Most people are interested in prolonging their (healthy) lives. At the moment the way we do this — reactively, dealing with acute situations — is expensive and far from ideal. We would all live longer, healthier lives if we could make smaller interventions when problems first become apparent through the data. And the NHS could probably be a lot cheaper.

Ultimately wearable technology stops being technology and it becomes fashion. Just like we expect fashion to incorporate technologies like zips and waterproofing, we will come to expect it to include digital technologies that capture information about us and our environments. There will be a few more flops along the way, but for me the rise of wearable technology is inevitable.

Posted by on

Project Santander Test Two: Controlling Sockets

WARNING: TECHIE POST. Following my visit to the smart city project in Santander with Telefonica last year, I was inspired to start building my own smart home based on similar technologies. This is partly an exercise of my rusty engineering skills but mostly about learning the realities of smart cities/smart homes through experience. I’ll write up the lessons in a much less techie form, but for those who are interested, I’ll also be documenting the detail here.

I’ve been carefully writing up my progress as I go with this project. Except for this post, the notes for which I somehow lost (these early experiments took place a couple of months ago). So here’s a description from memory of what I did and how it works now.

I want my Arduino-based network nodes to be able to trigger actions as well as collect data. That means turning on lights, remote controlling audio and video equipment etc. In order to do this, they will need RF and IR interfaces.

Using the same little cheapo RF chips I used in the first test, it is possible to interface with the HomeEasy range of sockets and switches, of which I have quite a few. Fortunately as with many Arduino projects, somebody else has done a lot of the hard work.

I used the HomeEasy Advanced library by Andrew Lindsay. Making this work first requires capturing the remote code from your existing remote. The Arduino then spoofs this. I had MASSIVE problems capturing the remote code and I can’t for the life of me remember why. Suffice to say that eventually it worked, and when it did, transmitting codes was really easy.

The trick will be finding a way to trigger the RF switches using the REST interface I’m planning to use for the network nodes…

Next: measuring power consumption via AlertMe.

Posted by on

Project Santander Test One: Capture Sensor Data and Send Wirelessly

WARNING: TECHIE POST. Following my visit to the smart city project in Santander with Telefonica last year, I was inspired to start building my own smart home based on similar technologies. This is partly an exercise of my rusty engineering skills but mostly about learning the realities of smart cities/smart homes through experience. I’ll write up the lessons in a much less techie form, but for those who are interested, I’ll also be documenting the detail here.

My smart home project will be based around a series of nodes in each area of the house. Each node will have an Arduino at its core and an array of sensors and other devices. Most of the nodes will have at least one analogue sensor and not all of them will be connected via a wired network. So I decided to combine two tests into the first experiment:

  1. Getting data from an analogue sensor in a usable format, and
  2. Sending that data wirelessly to another unit.

I warmed up by connecting up a sensor, in this case a photocell I had lying around, following the instructions here: http://learn.adafruit.com/photocells/using-a-photocell

Nice and simple.

Next I looked for some code examples using cheap RF units like these from eBay — again something I had bought on a whim because they were cheap: http://www.ebay.co.uk/itm/RF-Transmitter-Receiver-kit-433Mhz-wireless-Arduino-/321246183899

I found this: http://genericnerd.blogspot.co.uk/2012/07/arduino-sending-integers-over-rf-with.html?m=1

I downloaded the VirtualWire library and the code examples, and wired everything up. Note that the sketches don’t tell you which pins to use for transmitting and receiving — it’s 12 for transmitter data and 11 for receiver data. I uploaded the sketches to my two Arduinos and… nothing.

Actually not quite: the transmitter appeared to work fine, and when I uncommented the debug code I could see that the sensor was sensing fine too. So the issue appeared to be at the receiver end.

On second look I noticed that the transmitter was sending at 300bps and the receiver was set to 2000bps. There was also a configuration line that didn’t look necessary so I commented that out for good measure.

With both transmitter and receiver set to 300bps, I opened up the serial monitor on the receiver and…success! I was getting the sensor data through at the receiving Arduino.

This is a great proof of concept but it will need some work before it does everything I want. For example, I will probably be sending more than one dataset between nodes — e.g. light level, temperature, humidity, so I need a reliable way to identify what is being sent/received. I also want to make sure that the remote sensor batteries last a decent period so I’ll need to look at the intervals and the Arduino’s power modes. I know that the Jeelab libraries are good for this so may look at those down the line.

For now though, this will do enough to get me moving.

Next: RF remote control.

Posted by on

The Future of Giving: It’s The Thought That Counts

In December I spoke at the second Manchester Sunday Assembly, a gathering for people who recognise the great things a church can offer but aren’t so keen on the supernatural. If you haven’t been I can highly recommend it. There are groups popping up around the world where you can sing together, listen to interesting talks, drink tea and eat cake. There must have been over a hundred people at the one I attended at Manchester’s Victoria baths, a beautiful if chilly venue.

In the run up to Christmas the theme of the event was giving, so I decided to do a talk about the future of giving. I wanted to look at the way in which we give charitable donations, political support, and our own data might change in the future. This was a new area for me and not one in which I’ve done lots of research, so please forgive the use of anecdote and treat it more as an essay than an exercise in empiricism.

The slides on their own (embedded below and available at Slideshare) are a little incomprehensible so I’ve included the full script below too. This was written for me to say rather than be read as text, so please also forgive any grammatical errors.

 

Slide 1

A time for giving, a time for getting,
A time for forgiving, and for forgetting.
Christmas is love, Christmas is peace
A time for all hating and fighting to cease.

Slide 2

With those words, Sir Cliff blighted every Christmas album since 1988. But saccharine though they are, few can disagree with these sentiments. Christmas is a time for giving, and so I want to speak a little about giving in 2014 and how to be a better giver.

Slide 3

Human beings are good at giving. We’re a generous bunch as this chart from the CAF shows. If we have money, we donate it. While we may be feeling the pinch in Europe at the moment, the overall trend is one of growing wealth around the world. Millions more people are being lifted into the middle classes. Which means that millions more people will learn to play the game of psychological British bulldog we all play with chuggers on our high streets.

Slide 4

We like to think of ourselves as exceptionally generous in the UK, but the reality is that generosity is programmed into all human beings. Our own brains reward us for being generous in ultra pure, high grade narcotics.

When you give dopamine is released to the reward centres in the midbrain. And the subgenual area that is connected to social bonding is activated. Neuroscientists believe that years of experience has ingrained in our biochemistry the knowledge that we are much better off being part of a strong society. And so our brains reward us for contributing.

Generosity isn’t altruism, it’s science.

Slide 5

The problem is that when drugs are involved, natural or not, few of us make good decisions.

Which is why it’s so important to think about how, why and when you give.

Let me give you an example.

(If you don’t recognise him, this is Rob Ford, former mayor of Toronto. Google him if you don’t know his recent history).

Slide 6

The first thing I want to talk about is giving money to charity. This is a theoretical but it’s based on a real example relayed to me by someone working in the field.

Slide 7

Imagine you see a picture like this. You get a message: Help save the sight of this child for just £3. Text 7777.

A hundred thousand people click. The charity takes that money and uses it to build a clinic to provide medicine. Thousands of kids’ sight is saved. Great, right?

Now imagine the country they are targeting is failing and corrupt, but it still has the bones of a national health service. Its clinics are few and far between, and under resourced, but they are staffed by broadly skilled staff who treat a range of complaints.

A new crop of sight clinics pops up. Fresh, clean and widely publicised.

Now if you’re sick, which clinic are you going to go to? The fresh, clean one that is one hour’s walk away? Or the old decrepit one two hours walk away.

Unfortunately the first one only has the medicine and the mandate to treat you if you are under the age of 16 and have a specific sight condition. Having walked to one, you then have to walk to the other, angry and confused and that little bit sicker.

When you get there the other clinic remains overwhelmed and under resourced.

Now imagine the money raised here instead went to the national health service in that corrupt country. It doesn’t offer a compelling picture to put on the posters. Lots of the money might go into the pockets of corrupt officials. But the incremental improvement in that run down clinic might — in the long term — save many more lives and promote generally better standards of public health, driving economic as well as physical welfare.

Slide 8

Today we have no way to know if this is the case or not. But I’m told it’s the fear of many professionals working in international aid. What we need is better evidence to allow us, the donating public, to make better decisions based not on heart-rending pictures but based on genuine outcomes.

In the future we may find that the proliferation of mobile devices and social networks begins to provide us with this evidence.

Imagine if you could use social media to collect empirical evidence about the impact donations had on the ground in different countries. Small amounts of data from lots of people aggregated to create a clearer picture of aid need, outcomes, and the barriers that might prevent aid achieving its goal. Imagine if we could make our donations in the knowledge that every pound we put forward would have the maximum impact.

That might get our dopamine and oxycontin really flowing.

Slide 9

Now social networks and mobile devices are of course already deployed in social campaigning. Call it armchair activism. Call it slacktivism. Call it clicktivism. These days it’s very easy to lend your weight to an argument.

But should you?

This is the second type of future giving I want to talk about: giving your support to a campaign.

Slide 10

Take Avaaz, one of the largest networks for online activism. Avaaz isn’t all that popular with many traditional charities and campaigners. In part because it has achieved incredible scale and impact in a very short space of time. It has more than 30m members. But also because of the means it has used to achieve that scale.

Avaaz selects campaigns based not on a set of core beliefs, or even some empirical measure of which issues are the most important. It selects cases it believes will scale, and cases it can win.

Optimising campaigns for scale and impact often means dumbing down the argument. This is the one that caught my attention.

Now there is little doubt that the rush to first generation biofuels did drive food prices up. And that many first generation biofuels were less environmentally friendly than petrol.

But without subsidies there is little incentive to develop the next generation of biofuels. Ones that place less strain on food supplies and the environment. Ones that truly can be a weapon in the fight against climate change.

Dumb down your argument too far and suddenly you find yourself on the same side as big oil and climate change deniers.

Now it is hard to generate mass support for a campaign that gets into the nuances of subsidy policy and science. But without doing so campaigning is a blunt instrument that won’t always steer debates towards the best outcomes. Just those that look best when only considered in black and white.

Slide 11

Now just as it as first generation biofuels that it was campaigning against, Avaaz is itself a first generation online campaigning platform. I believe it will improve as the older campaigning organisations, currently complaining about its success, begin to adopt some of its tools. In fact to be fair to Avaaz it did improve this campaign, making it much more nuanced as you can see here.

What Avaaz has proven is the power of collecting users — and their data — online and directing them at specific causes. Which brings me to the third type of giving: data.

Slide 12

We have become very used to the value exchange that happens online. Hand over your data in return for a piece of content, or access to a social network. We’re comfortable with that because we get value back. It feels free to us because we have no way to sell the data we’re currently trading for access to a social network.

Slide 13

But if the data we’re giving up is really of such low value, then why is Facebook worth over a hundred billion dollars? Why in fact is the value of each of the major social networks roughly £60 per user?

Slide 14

If I go out to buy some consumer data I’ll expect to pay between 60p and £6 depending on what level of information I’m getting. Per thousand records.

Now the discrepancy between that and Facebook’s value is in part down to repeat sales: it can effectively sell our data over and over again. But it takes a lot of 0.0006 pounds to get to £60.

They’d have to sell access to every one of us a hundred thousand times.

What makes the difference is intention: social networks have so much information about us that they can put ads in front of us that we are more likely to click on. And as soon as we click, we are adding something vital to our data: intention.

Once you add intention to data, values go through the roof.

So, you’re advertising private health care on Facebook. Facebook knows their age and lifestyle. Facebook places adverts in front of the people most likely to click on them. Now I haven’t checked the cost per click for private health care on Facebook. But on Google, it’s £30 per click.

Suddenly Facebook only has to sell two clicks per person to justify its valuation.

Slide 15

Forgive my assumptions, but I doubt many people in this room are shopping for private health care. But we all shop. And increasingly we find our purchases online. Social networks make a lot of money from our data by using it to sell to us.

Referral schemes like Quidco are already helping us to share in the value of this data by sharing the fees they get for introducing us to retailers. But this is a clunky process at the moment. Increasingly I think we will take control of our own data and expect to be rewarded for sharing it with more than access to a social network.

Slide 16

In summary, when we’re giving money, we need to think more about the outcomes and less about the activities we’re funding. Today that is hard because we have little direct connection to those experiencing the outcomes. But Facebook has a staggering growth rate in the developing world. Mobile devices are proliferating at an incredible rate. What we need is an evidence network that enables us to rapidly collect and analyse data to make better decisions.

When we’re giving our support, we need to be more discerning. At the moment complex arguments lose out to simplified effective rhetoric — as has always been true in human history. The opportunity of technology is to better communicate the complex, but we need to think more too. Just like most of us have learned that those pages on Facebook that invite you to click if your mum is great or you have blue eyes are basically scams, so we must begin to differentiate the worthy causes from the link bait.

On giving our data, we need to value ourselves more. We need the tools to help us measure our value to advertisers and a brokering system so that we can access that value.

In short, whatever you are giving in 2014 remember: it’s the thought that counts.

Posted by on

Project Santander: How a Smart City Inspired a Smart Home Project

In October I visited Santander, Europe’s most advanced smart city laboratory, courtesy of the project lead Telefonica Digital. The city has a grid of 12,000 sensors feeding the local government information about energy consumption, traffic, noise pollution, air quality and more, enabling it to make better decisions and reduce waste. At least that’s the theory: this giant experiment is still active and hasn’t yet delivered measurable, reportable results.

Nonetheless I was impressed with what I saw and it inspired me to pick up my own home automation efforts. Hence the name of this project: Santander.

My plan with this project is to develop a simple, cheap, modular network node that can feed information back to a central control and information core. The node should also be able to receive instructions from this core and relay them out to local devices. I want the device to be modular because different rooms have different sensing and control requirements. And I want its network interface to be fairly standard so that it can be tested with different core platforms — there are loads of options out there.

I’ve played with LinuxMCE before and frankly got a bit lost. This time around I’m looking at MisterhouseOpenHab, and FreeDomotic. All of these are based on languages (Perl, Java) with which I have no experience. Eek.

Since the software side of this project is a little scary, let’s start with hardware.

Just like the academics from the University of Cantabria who have built much of Santander’s sensor network, I’m starting with Arduino. It’s cheap, flexible, extensible and easy to program with lots of libraries and code examples available.

Onto my Arduino I’m initially going to add two shields: one Ethernet shield, and one Sensor shield. I’ve chosen wired Ethernet as my primary means of connection because it is cheap, reliable and not battery hungry. Plus I already have three rooms in my house wired up.

The sensor shield will allow me to plug in various different sensors for light level, temperature, humidity, and presence, or local interfaces such as IR or RF, in a slightly more reliable/robust way than jumper cables, and in a more reversible way than soldering them onto a proto shield.

Because it seems to be the most common standard for interfacing between sensors and hub software I’ll primarily be working on a RESTful HTTP interface. But I’ll also experiment with off-the-shelf software for the Arduino like Souliss.

So, I already have a couple of Arduino boards to play with, and a sensor shield, some IR LEDs (harvested from old remote controls), and a 433MHz transmitter and receiver module. I’ve ordered a PIR sensor, a combined temperature and humidity sensor, and an Ethernet shield.

My first experiment will be taking readings from a light sensor — a simple photocell. Then I’ll try sending and receiving HomeEasy commands using the RF transmitter receiver. I’ll also try communicating between Arduinos with the RF transceiver pair for those sensors that will be away from power and wired Ethernet. IR controls will be another test before all the ordered components arrive and I can start to bring it all together in a single package and sketch.

Finally for this scoping post, energy. I run an AlertMe system left over from an old review to keep an eye on my electricity usage. I plan to plumb this into whatever hub software I choose to use rather than adding another means of electricity monitoring. It has an API and I’ve already had a little play with accessing it via a PHP library. This doesn’t do anything for my gas consumption though: I’ll need to find a way to get a reading off this.

Onwards and upwards…

Tom Cheesewright