A few weeks ago, the Financial Times Digital Business podcast visited Hursley and checked out some of the innovations that are being worked on. The result is a nice 22 minute episode which tours the lab (including the Retail Lab pictured on the left!) and talks to John McLean, Andy Stanford-Clark, Bharat Bedi and Jamie Caffrey.

If you’re into augmented apps, location awareness, Emotiv headsets (as featured in our last post here, too!), e-paper labels on shop shelves, telemetry, instrumented houses, and Smarter Planet – it’s a great listen.

If you watched this week’s edition of Bang Goes The Theory on BBC1, you will have seen Nick O’Leary and Kevin Brown from IBM Hursley helping Jem and Dallas to drive taxis. That probably wouldn’t have been entirely revolutionary, had it not been done through a combination of an Emotiv brain-signal-reading headset, and some MQTT and Arduino funkiness… no hands on the wheel or feet on the pedals!

Nick has a great write-up of what sounds like a fun (but cold) event. You may still be able to catch the fun on iPlayer, or there are some clips over here.

(Image: Creative Commons Attribution Non-Commercial Share-Alike (2.0) from knolleary’s photostream, used with permission – full set)

There’s a snappy new website about IBM’s Ideas Lab in the UK. It talks about the kind of people who work here, and has links to a virtual tour of the site, a Smarter Planet video, blogs by recent graduates Ed and Clare, a simulation game, and some pictures of shoes.

So where are all those brilliant biologists, fantastic philosophers, genius geographers, clever chemists, marvellous mathematicians…? You’ll have to come up with your own descriptions for English, History, Psychology… but they should apply, too.

[oi! I'm an historian thanks very much Peter! ... hmm... enlightened English graduates? psuper psychologists? hintelligent historians...? - Andy ... and if you want to read more about what life is like at Hursley, there was a lovely piece in the Sunday Times "Inside the IBM Dream Factory" a couple of weekends back]

You can also browse the set at Flickr.

It’s like a glimpse into the computing past. I’m a particular fan of the futuristic “Mind Appliance” (looked and sounded great, but turned out to be filled with BASIC programs), and the book about the exciting new computer language, PASCAL.

To be fair though, I did end up picking up a free copy of The 7 Habits of Highly Effective People, which Martin has been recommending to me for ages.

Andy Stanford-Clark hooked up a set of lights to Twitter. As reported in Computer Weekly:

Using some clever IBM middeware, The microcontroller sets the illumination colour based on a signal from the internet or via SMS over a GSM network – so you can tweet “ibmlights” with the word RED, GREEN or BLUE to change their colour.

In fact, the commands got a bit more sophisticated than that, with more colours and lighting patterns. Towards the middle of last week the lights ended up over Laura‘s desk, and a growing band of folks delightedly tweeted the @ibmlights account with instructions to change colour or pattern. She took some pictures for me (and some video as well, but I didn’t have time to edit it…).

Another year of innovation and fun at Hursley! :-) Happy Christmas!

(by the way, well worth taking a look at the rest of the Computer Weekly article I linked above – lots more coolness from Hursley! oh, and I’m not sure how long the lights will be online… it’s just a bit of fun really)

A few people have asked “why Wolverhampton?”. Which is a fair question considering that four of the organisers live in Hampshire, one in the South-West, one in Liverpool, and one in Bonn.

Well, Wolverhampton is the location of the annual LugRadio Live open source community conference. The organisers of LugRadio Live (the LugRadio podcast presenters) are, or were, based in Wolverhampton. While there are many things you could say about Wolverhampton, one thing that always impressed me was that, to attend LugRadio Live, people flew to Wolverhampton from all over the UK, from all over Europe, all over the States, and even from Hong Kong and Australia at times (see my blog post about past LRLs for more).

Last year, after four hugely popular LugRadio Live events, including one in San Francisco sponsored by Google, the team announced that the fortnightly LugRadio podcast was going to end, and so the fifth LugRadio Live (in July 2008) would be the last ever LugRadio Live.

And then, under pressure from Popular Demand, they agreed to do another last ever LugRadio Live – in October 2009. This last ever LugRadio Live, though, would only be one day, the Saturday, like their first ever LugRadio Live. Which left a whole Sunday to fill. Which is where OggCamp comes in.

When we decided to organise OggCamp, we had no idea how it would go down. We figured that, between the two podcasts (Ubuntu-UK Podcast and Linux Outlaws) we’d had enough positive feedback that we could get at least 50 people along. Because it would be the day after LRL, there was a chance that enough LRL attendees would stick around for the day on Sunday and coming to OggCamp too. To make extra sure of that, we decided to hold OggCamp in the official LRL hotel (so that the geeks could just roll out of bed and into OggCamp), and to make the event free to attend.

In the end, about 130 people came to OggCamp. Which was brilliant!

The sight of people queuing up three flights of stairs to come in at 10.30 on the Sunday morning left us briefly gob-smacked.

We kicked off at about 11am with a quick introduction from all the presenters in which we explained how there was no pre-arranged schedule and that to sign up for a talk you just had to write it on a sticky note (large notes for full-hour talks; half-sized notes for half-hour talks) and stick it in a slot in the grid on the wall.

First up was Andy Stanford-Clark who did a brand new talk, specially written for OggCamp (and completed the night before while the rest of us were at the LRL kareoke party), about the geekier details of his Twittering House (the stuff the BBC didn’t get!). By midday, the schedule was getting pretty full (something of a relief!) and the planned topics included web services, how to prove identity on the Net, how to encourage young people to use Open Source Software, politics and geeks (from ORG), translating Playstation 2 games, and how to explain programming to non-programmers!

At 3pm, everyone gathered in the main room to watch a live joint recording of the Ubuntu-UK Podcast and Linux Outlaws. This started with a live raffle draw (surely a first in open source events?) for some very cool prizes donated by our sponsors, including a couple of Viglen MPC-Ls, some Ubuntu laptop bags and hoodies, an O’Reilly book, and an Arduino Mega. After the raffle, we did two segments: one about producing media using Open Source Software, and one about whether or not the Open Source community spreads itself too thin by creating so many different distributions. The segments included a lot of audience interaction, and also real-time twittering from the audience on to the TwitterFall screen behind us on-stage.

The live show was something that we had been nervous about because six is a large number of people to be talking anyway but also because the two podcasts (UUPC and LO) are quite different in style so we had no idea how well we would integrate. The two podcasts released their own versions of the live show during the following week and, if you’re keen, you can compare and contrast the two: UUPC (family friendly) and LO (includes the naughty words). I don’t think either podcast did much editing of content, which drew this comment from a UUPC listener.

So, all in all, I think we can say that OggCamp was a success. :)

It was certainly a lot of fun – if exhausting!

We also sold enough raffle tickets and OggCamp limited edition souvenir mugs to financially break even on the whole event. Which was good from our point of view. And there has been a load of positive feedback from the attendees, including questions about whether we’ll do it again next year. Although we’ve tried to not to commit to anything, by now I think we can safely say that there is likely to be another OggCamp next year.

(For more photos, see the OggCamp group on Flickr.)

The theme for the day was location and augmented reality.

A particular highlight was a talk by Paul Golding on Augmented Reality & Augmented Virtuality, covering a variety of topics such as the state of Virtual Worlds today, and the potential of mobile augmented reality apps to move us from a “Thumb Culture” to a camera-led “Third Eye culture”.

A number of mobile augmented reality platforms were discussed, such as Nokia’s MARA research project, the QR-based Insqribe, the real-world / virtual-world mobile mashup platform junaio, and the ‘world browser’ Wikitude.

Another platform that got several mentions, including a developer’s crash course in the afternoon from Richard Spence, was Layar.

I had a quiet afternoon in the office this afternoon, so I thought I’d give the Layar API a quick try for myself.

Layar is a mobile app for Android and iPhone that lets you display location-based information overlaid on a real-time camera view.

For example, the screen normally shows a viewfinder-like view from your mobile’s camera.

Search for “coffee” and a bunch of markers appear on the view, showing you where the nearest coffee shops are.

As you move the phone around, the markers follow the approximate location of the places they are showing you.

That’s assuming you want to search ‘Google Local’, but that’s not the only option. Location data is provided through “layars”, and there are layars available for location-tagged Wikipedia articles, Flickr photos, brightkite users, and more.

The interesting thing talked about at openMIC was the Layar API which lets anyone create a new Layar with their own information.

So I decided to spend a quiet Friday afternoon in the office creating a Layar for around Hursley. :-)

This means a phone with the Layar browser installed can browse and search for points of interest around the site.

It was really very easy, so I’ll quickly outline the steps involved.

Step 1 – Get an API key

Visit dev.layar.com and click on ‘Request a developer account’.

Once approved (it took a few hours), you get a developer ID and key.

Step 2 – Define a new layar

Once you’ve got an account, go back to dev.layar.com and fill out the ‘Create Layer’ form, describing the layar you want to create.

There are a ton of options here. These range from the obvious like a name, description and tags to display to users choosing a layar to use, to neat customisations like uploading custom colour schemes and icons to use as map markers on the phone screen.

I just filled in a brief name and description and left the rest as the default values.

Step 3 – Define a source of Points of Interest (POI)

Layar wants to support real-time location information.

A good example of this is the brightkite layar. Users choosing the brightkite layar see a view overlaid with the current location of their friends or other brightkite users.

To make this possible, you don’t create a Layar by uploading a static database of points of interest.

Instead, you have to provide a web service that Layar can query to get current locations. (When defining the new layar on the dev website, one of the steps is to provide the URL of where you will put this web service.)

The API doc for this web service is at layar.pbworks.com and details the parameters that your web service will be invoked with (e.g. the user’s current location), and the format for the response that your web service must return to Layar.

The Layar browser does perform client-side filtering to only show markers relevant to the user’s location. However, if you are writing a Layar like the Wikipedia one, it’d be best not to just return your entire database of POIs, burning the mobile’s battery both in the data transfer and the processing needed to do the filtering.

For my quick first test, I was a bit lazy and ignored the input location parameter and just return all of my Hursley POI markers regardless, leaving the mobile to do the filtering.

Parameters are provided as URL variables, and the response format is a pretty straightforward JSON interface.

It took no time at all to knock up a quick bit of PHP to return this…

$currentLat = $_GET['lat'];
$currentLon = $_GET['lon'];

$place1 = array('actions'=>array(array('uri'=>'http://maps.hursley.ibm.com/showmap.html?q=reception',
                                       'label'=>'view on map'),
                                 array('uri'=>'tel:01962999999',
                                       'label'=>'phone now')),
                'id'=>'00000001',
                'title'=>'Main Reception',
                'imageURL'=>null,
                'lat'=>51026570,
                'lon'=>-1397964,
                'line2'=>'A1101',
                'line3'=>'',
                'line4'=>'',
                'attribution'=>'ETS demo',
                'distance'=>distance($currentLat, $currentLon,
                                     51.026570, -1.397964),
                'type'=>0);

// <snip> ... bunch of other places here ... </snip>

$allPlaces = array('errorCode'=>0,
                   'errorString'=>'ok',
                   'hotspots'=>array($place1 ,$place2, $place3,
                                     $place4, $place5, $place6,
                                     $place7, $place8, $place9),
                   'layer'=>'ibmhursley',
                   'morePages'=>false,
                   'nextPageKey'=>null);

echo json_encode($allPlaces);

Obviously, hard-coding all of the locations is a sloppy way of doing this – in a real system, you’d want the PHP to be retrieving this from a data store of some sort.

There is open source code available from a bunch of places to get you started in doing this – PorPOISe looks like an interesting example.

Interestingly, the mobile client doesn’t calculate the distance from the user to the POI – you have to return that from your web service call. This is the only reason why I didn’t just write a static text file with JSON data for my “web service”, as I had that one dynamic bit needing to be changed.

The other bit worth highlighting is that each POI can include one of more “actions”, that get offered to the user if they click on a point of interest on their mobile. I added a few examples such as launching a web page or making a phone call. E.g. you can see where Main Reception is on your screen, and choose to call Reception if you want.

Step 4 – Test the web service

It took a couple of tries to get my PHP right… mainly because several of the parameters described as optional in the API doc are actually required :-)

Fortunately, there is a nice test service available at dev.layar.com/api20test.

You plug in your developer ID and key, give it a location of where your mobile should be, and click “Load POIs”.

It tells you what is wrong or missing with the JSON response it gets back from your web service, which for me was either which required value was missing, or a numeric value that I was returning as a string. The errors are clear enough to help you fix your service.

Step 5 – Try it for real

layar.pbworks.com has a link to the installer for the Android app.

This app is a developer version of the normal one available in the Android Market.

(If you already have the normal app installed, you need to uninstall it first. I didn’t, and the developer app didn’t work properly until I uninstalled and started again).

The developer app gives you extra options in Settings to provide your developer ID and key. This lets you access your unpublished layar for testing (until approved by Layar, your new layar isn’t available to other users).

(It also lets you override your phone’s current location – useful for testing indoors!)

Once you’ve done that, your new layar shows up in the mobile’s list of “Featured” layars.

(I found that it only showed up if I also changed the Country setting from “Auto” to “GB”… not sure why, although when I defined the Layar, I did specify that it was a GB layer.)

And that’s it… I had a fun excuse to go wandering around Hursley this afternoon to test it out and collect the screenshots for this post. :-)

Hursley is a culturally as well as a technically diverse place, so we’ve got some great opportunities to learn from each other. This lunchtime I popped along to one of the events organised by the lab’s Chinese Connect team, which was all about Understanding the Chinese Calendar (the title of the post refers to a significant date this week in that calendar, September 23rd).

Previous talks in the Chinese Culture series, which is organised by Hursley’s Jenny He, have covered subjects such as the evolution of the Chinese languages, how to understand Chinese names, and Chinese music and instruments. I’m embarrassed to say that this is the first of the talks I’ve been to, despite working here for some time… I really should take more advantage of the range of activities and opportunities that Hursley has to offer!

Today’s talk was delivered by Darren Beard, who was particularly interested in the astronomical background to the Chinese calendar (having published a paper on the same topic several years ago). Darren covered the scientific background of this lunisolar calendar, and the changes that have taken place to it historically over the ~3500 years it has been around – particularly interesting to me, since I’m a historian by background. It’s a complicated system which takes account of 19 year lunar cycles, requires things like leap months, and has a set of rules which specify how it works… but it is certainly more comprehensible once you understand those aspects. It was interesting to realise just how much my own perceptions of time are based on the calendar system I’ve grown up with!

As is usually the case when I’m attending a HantsLUG (Hampshire Linux User Group) meeting, it was a lovely sunny day on Saturday. It’s like as if it knows that I’m going to be spending the day inside, geeking in front of a laptop screen. This meeting, however, we put the sun to good use, first of all showing off Hursley Park at its best, and then lunching out on the decking at the Clubhouse.

When I was a more frequent attendee of the HantsLUG bring-a-box meetings (where I installed my first Debian distro, and later my first Ubuntu), I’d often thought how cool it would be to host a meeting at IBM Hursley. But I never got as far as investigating the security and wifi hassles I’d have to overcome. Fortunately, Anton Piatek was a little braver and sent some emails to nearly the right people (who helpfully forwarded them on to really the right people), and suggested his plan to Adam Trickett, Chair of HantsLUG. Adam says he nearly bit Anton’s hand off and so it happened.

HantsLUG is one of the biggest LUGs in the UK and is our local Linux user group but has surprisingly never really (in the 7 years I’ve known them) had a huge amount of interaction with IBM Hursley. For a long time, though, there has been a good pool of Linux skills and interest in the Lab, and over the last couple of years the number of people around the Lab voluntarily using Linux as their desktop OS has risen (as has the number of Ubuntu lanyards to be seen as you walk the corridors of Hursley).

So what makes Hursley a good place for a LUG meeting? Well, for a start, it’s just a really nice place to be – and Hursley House as well as the Park are very impressive to show off to visitors :-)

On Saturday, we were mostly in the Auditorium (where Spitfires were built during WWII), then when we led everyone down to the Clubhouse for lunch, we took the usual site tour scenic route via the Sunken Garden and fish pond. Although Hursley is out in the country, seemingly the middle of nowhere, it’s actually on the bus-route from Winchester so we had an excellent turnout of about 30 people. IBM Hursley also has a lot of cool people who do cool things that we can tell people about (although one piece of feedback I heard from a LUG person was that they thought we didn’t talk enough about what IBM does!).

Although we had the House to ourselves, and everyone was free to stand around and chat in the Main Hall, most of the day revolved around talks in the Auditorium. It all kicked off at 11am with an introduction to IBM Hursley (and, of course, directions to the fire exits and toilets) from Anton. The inimitable Andy Stanford-Clark, fresh from a week of press interviews, enthused everyone till lunchtime with tales of mouse traps, MQTT, twittering houses, twittering ferries, water meters, and energy monitoring. I say ‘enthused’ but there must be a better term to describe the way the audience rushed the stage when Andy offered to sell Current Cost monitors at a discount…

After lunch, we had a collection of shorter talks on a range of topics:

• I talked about InfoSlicer, the open source software that my Extreme Blue student team developed last Summer and IBM released under the GPL
• Anton described the anatomy of Ubuntu packages (he’s the guy that provides Ubuntu users in IBM with the flawless packages we’ve come to rely on)
• Tony Whitmore related his experiences of producing the popular Ubuntu UK Podcast – and pimped the upcoming OggCamp unconference
• Adam Trickett, Chair of HantsLUG, gave out free books in return for promises of book reviews on the HantsLUG wiki

Then everyone just hung around chatting for ages.

It was a really enjoyable and relaxed day; kudos to Anton, Stephen, and John for organising it from the IBM end. Thanks also to the IBMers who came along and to the many HantsLUG members who turned up. I’d say it was a success and we should definitely do it again.

I believe it is important to leave it entirely up to the individual to recognise the positives of IBMers they work with, and go from there – you can’t do any better than this, but, there’s a bonus: sometimes IBM wants to recognise you too!  To make it all look good, to customers, or to make people happy, and to maintain the technical and innovative reputation or brand IBM has, they’ve built title names or award names as follows:

• IBM Fellows
• Corporate and Patent Portfolio Awards
• Distinguished Engineers
• Members of the IBM Academy of Technology
• External Honours
• Major Outstanding Innovation and Major Outstanding Technical Achievement Awards

Lots of different names, but of course IBM is so diverse that it’s difficult to recognise things in a systematic manner. Looking at the titles, I think the key points here are: if you love corporate technology and/or innovation, you can’t go any better!

I did.  As a result, I fell under the External Honours category, primarily for the RADAR Young Person of the Year award that I received last year.  Thereby I got into the book and, more excitedly, to meet other people and learn more about IBM.  What did I learn from the other IBM attendees?

I would firstly ask where do you work?  If it’s GBS [IBM Global Business Services], I would ask if they’ve heard of SWG [IBM Software Group]?  If yes, have they worked with SWG before?  With either answer, I would then ask if they’ve heard of Lab Services [which is where I work].  With these answers, I started coming up with questions I’d love answers to – for example:

• how to improve the awareness of what Lab Services can do, across into GBS?
• does the linkage between Lab Services and GBS have any room for improvement?
• have people from Lab Services moved to GBS?
• is Research well connected with UK?  With Research being in Switzerland?
• is the market for deaf-related and/or blind-related technology too small to be of interest?

I also met with Brendon Riley [the IBM UK General Manager], who I was very keen to talk with, particularly as my wife is also Australian.  I was thrilled when he told me that he was from Perth – where my wife was from! Brendon very kindly took the opportunity to ask me to do something for him – clearly he was keen to demonstrate that the planet was flatter, as he could ask me directly like a colleague sitting at the next desk might, rather than passing the request down through a hierarchical organisation. The fact that he’s from Perth, my wife’s from Perth, the Country General Manager asking me directly to do something for him, as our CEO would say: the world is becoming smaller and flatter!

Ben