Kevin X Brown

"I'm a bus" Part 1 - The idea

2011-12-19T08:43:00.000-08:00

I'm working on an exciting little project at the moment, with some colleagues of mine from Hursley, which could have an impact on how people travel to work...

2500+ people work at IBM Hursley and so there is a minibus which picks people up from nearby Chandler's Ford and takes them to work three times each morning, and returns them at night. A while ago we instrumented the bus with a GPS tracking device and a button which allowed people to see its whereabouts using a map on their smart phone, and see how many free seats there were on board.

This greatly helps reduce the number of people travelling in their cars and back every day. But this is just one route - it would be impossible to provide a minibus from all the many locations that people come from to reach Hursley! But what if the cars coming in and out of Hursley were also able to act as minibuses, and pick up other IBMers along the way? Then, we would have mini-minibuses coming from Winchester, Portsmouth, Alton, and many other places!

So the idea of the "I'm a bus" scheme is that car drivers using a smart phone can record their journey to and from work, and upload it as a bus route on the "I'm a Bus" website. Every day, when they leave home, their route becomes active, showing the car's current location. Passengers who need a lift in to work can stand in a safe place on the route and request a lift, via their own phone. The driver's smart-phone will announce using voice technology that "Joe Bloggs has requested a lift in 1000 meters on Hursley Road" and will display a picture of them from our internal employee directory, so they can be recognised. The shared journey will be recorded, and you can see the environmental savings you and everyone else combined (in terms of miles saved and carbon dioxide) has made on the website.

This means people can measure their contribution to lowering greenhouse gas emissions, develop a wider network as they pick up new faces en route to work, and also, as the driver, you can specify a preferred contribution, which the passenger will see prior to requesting a lift. This might be a "cup of coffee", or a suggested contribution per mile.

Of course, once this is running, there are lots of enhancements, which could be made, including allowing traffic / weather information to be shared from web feeds or from other driver's information, sharing planned routes as well as impromptu routes, and provide a full list of possible options for getting around (eg ties with bus, train and taxi services)

So.... this is the *idea*. Turning it into a working system is something we've been working on for a while. It will rely on good intuitive technology that doesn't require lots of learning, and it will require enough people to use it, to make it effective. In my next post, I will reveal how we're getting on!

Kinect Libraries

2011-04-22T05:59:00.000-07:00

Today I've been looking around at the libraries available for Kinect development. There's lots of hacking going on and so I wanted to create a list of libraries so as not to get confused. I've made it public in case it's of use to others (and feel free to correct me if I have made an error).

OpenKinect http://openkinect.org
This is the main page for the open kinect libraries for Linux, OSX and Windows. There is a highlevel API and wrappers for Python, C, ActionScript, C++, C#, Java, Javascript and Lisp.

depthJS http://depthjs.media.mit.edu/
An extention for Chrome/Safari from MIT media lab that allows javascript access to the kinect. Unfortunately only the source is available and it's not a simple install at the moment, but looks promising.

SensorKinect http://www.primesense.com/ https://github.com/avin2/SensorKinect
PrimeSense are a company that make the PrimeSensor, a product on which the Kinect was based. They have drivers to convert real images to depth maps (called PrimeSense/Sensor). However they have also created a driver for the Kinect, called SensorKinect.

OpenNI - http://openni.org
OpenNI supplies a set of APIs to be implemented by the sensor devices, and a set of APIs to be implemented by the middleware components. A very interesting PDF is the OpenNI Userguide.

NITE - http://www.primesense.com/?p=515
This is "middleware" made by PrimeSense, which plugs into the OpenNI framework

FAAST - http://projects.ict.usc.edu/mxr/faast/
The Flexible Action and Articulated Skeleton Toolkit sits on top of NITE, and provides tracking of whole skeletons. (Windows only)

openNI Kinect http://www.ros.org/wiki/openni_kinect
This focusses on the integration of the Kinect with ROS (Robot OS)

I will keep updating as I discover more!

Emotiv Headset and Locked In Syndrome

2010-10-18T05:53:00.000-07:00

In March, a salesman working for IBM had a stroke, which left him with complete paralysis, unable to use his muscles, and without the ability to speak. His brain however is working fine - a condition called Locked-In Syndrome. His means of communicating is by his eyes - looking up for yes, and down for no. He has to wait for someone to ask him if he'd like to speak before being able to do so. Then, using a letter chart, that someone must point at letters one by one, until a confirmation is received. This is the first letter of the sentence. The process must be repeated for every other letter in the sentence, until the full sentence is spelt out.

The salesman in question is called Shah, and Sarah (my wife) is his Occupational Therapist (OT), who saw the mind-reading headset from Emotiv that I was using in IBM and thought about the potential this could have for Shah. The headset was designed for the gaming industry, and measures facial expressions, excitement/boredom levels and can be trained to listen for particular thoughts, which can perform an action on the computer.

Sarah asked me to demonstrate it to the consultant, speech therapist, psychologist and hospital ward manager - who agreed it may have some potential, and who took it to the ethics committee. Shah was informed about the headset and what it can do - but that we weren't aware of any person with the symptom who had tried it before. Shah, being a bit of a techy, was up for pioneering it - so I met him last week for his first go.

Just like the many other people I have demonstrated the headset with, Shah was instantly able to train the system so that one action - the "push" action - would push the floating "Emotiv cube" into the distance. As we trained, the ability to push (and stop pushing) at will, improved.

Adding a second action adds a complication. Now the unit has to distinguish not just whether you think or don't think, but which thought you are having - much harder! And like everyone, it will take a bit of time to practise getting sufficiently good at this to control. However Shah is up for the challenge, and has a fantastically supportive family who are willing to help him train.

Last week, after one week of use, Shah had managed to train the headset very well - though when tired the Emotiv skill can go down as well as up. Working with the speech therapist, we have connected the output of the headset to the input of "The Grid 2" - a piece of software by Sensory Software which allows a user, normally via eye tracking, keyboard, mouse or switches, to control their environment, write emails and surf the internet. We have initially set up 3 different menu items, and thinking "push" starts highlighted these options in turn at 10 second intervals. Thinking push again, will select one of the options. Sounds easy perhaps - but if I tell you to *not* think of a rabbit - you can't help but think of a rabbit! It's hard to think about when to start and stop thinking, and then change to think about the thought you are supposed to be thinking about! I think this will take some practice to achieve, but once achieved could be widened to many more options, with very little extra practice.

Often people ask about using the headset for people with different conditions, and in different places. What with Shah being a fellow IBMer who likes gadgets, and his OT being my wife, and with the particular medical condition he had, conditions were ideal to try out the headset. It will take time for us all to get it right - but for now it is looking good. I will try to keep you updated!

Wearable computing devices

2010-05-27T04:30:00.000-07:00

Time to do a bit of blogging again. I've recently been asked about various wearable computing devices... here's a starter list, but I'd be interested in any other links you may have.

AR Glasses: These plug into a phone and can overlay data like directions / emails etc whilst travelling.

Gesture Gloves: Very accurate hand movement recognition.

Gesture Watch: Accelerometer, temperature, pressure sensors - also can act as wireless hub for other sensors.

Sixth Sense (wearable video/projector): Gesture recognition, ability to detect objects and augment real objects with data by projecting back on to them. Very cool.

Emotiv Headset: Brain reading device built for gamers which reads facial expression, excitement/engagement, trained actions and head movement.

NeuroSky Mindset: Brain reading device which measures excitement/engagement. Chip has the ability to add other sensors.

Mobile Phone Jewellery: Article from 10 years ago about IBM going into jewellery that can be used as a mobile phone.

We have used some of these sensors in the ETS Lab... more in a follow up post!

X200 Laptop from IBM On Demand Community

2009-02-14T09:55:00.000-08:00

IBM set up the On Demand Community programme some years ago, which provides money or IBM products to eligible community organisations and schools where IBM employees and retirees are actively volunteering (measured by recording hours in an online recording system), and in support of specific projects.

Four years ago, the 1st Chandler's Ford BB DofE Group applied for some funding for training, and received over £500. Training does not come cheap, with some Walking Group Leader courses running at around £400 per person, and the Assessment at a similar cost. Our First Aid courses are run by St. Johns Ambulance in Eastleigh, who help us by letting us donate a relatively small amount for a 2 day course. Leaders generally have to pay for this themselves, so it was good that, thanks to the contribution from IBM, the group could subsidise some of this training.

At the end of 2008 we applied for an IBM technology grant, and were very happy to be awarded a new X200 laptop. Using this, we will be able to maintain records of participants, leaders, training materials, and other resources needed to run the group - as well as using it for route planning using MemoryMap and helping young people access the DofE website if they need to get advice on completing their Skills, Volunteering and Physical sections of the Award.

We run the Award by charging participants as little as possible in order to prevent young people from missing out on opportunities due to the cost. This means however, we are completely reliant on donations to "enhance" the group in ways that this donation from IBM has, and will do. The BB owns one laptop which we can borrow for route planning, but with 6 groups attempting to route plan using the same computer, it makes a big difference to have our own! So a public thanks again to IBM for this donation.

Using Brain Computer Interfaces for something useful

2009-01-13T14:54:00.000-08:00

There are now a whole collection of Brain Computer Interfaces around... and this latest one from the games company Mattel uses the power of concentration to make a ping pong ball hover. The technology is supplied by NeuroSky, who have also provided the technology for a Star Wars game - $50-$100 for the fun of using "brain waves to allow players to manipulate a sphere within a clear 10-inch-tall training tower..." - something I could imagine myself doing just the once. Never-the-less it's quite good to see the technology coming down in price.

What fascinates me however is how we might be able to use brain waves for real personal benefit and integrate it more with our every day computer interaction experience. People who have severe motor difficulties could use it as a rehabilitative aid. Performing product research, it can be used for detecting when people are excited or not when they experience a particular product or website. When we are browsing on-line, we could combine it with eye tracking to provide feedback on what really interests us, and then tailor our online experience accordingly. When we go to the shops... the cinema... driving in the car... feedback can be provided to enterprise systems in all sorts of situations, which in turn can affect the environment around us. I'm really looking forward to talking with a number of customers about how they can integrate this and other interface technology in their own innovative projects in 2009!

Calculating energy usage is harder than it seems

2009-01-06T08:59:00.000-08:00

In my last post, I said a Christmas project was to monitor my usage - but I confess I've got a problem of how to measure my energy usage which is so far unsolved.

As I said, I have a currentcost meter (m1) which reads the current going into and out of the house. Of course, it cannot determine which way current flows, so it reads 50W regardless of whether the 50W is imported to, or exported from, my house. Therefore, I put a second meter (m2) round the wire from the solar panel, which told me how much the panel was generating and was thinking that if the value on m2 > m1 I must be exporting, otherwise I would be importing. However this thinking is flawed:

If I am generating 100W, (m2=100) and I am using 50W, then I will export 50W (so m1=50).
If I am generating 100W, (m2=100) but I am using 150W, then I will import 50W (so m1=50).

So from my two readings, I cannot conclusively say whether I am importing or exporting electricity, unless I am missing something obvious? The bright green import/export meter which Southern Electric fitted manages to work out what is being imported and what is being exported - but there doesn't seem to be a nice little serial port anywhere obvious! The other option is to buy a wireless unit as an addition to the inverter but this costs £120 and doesn't appear to give me open access to the data.

BTW - in 20 days we've generated 16.2 kWh and exported 9 of them. A grey day typically generates 200Wh the whole day - a sunny winter's day like today seems to generate about 2kWh.

The solar works

2008-12-18T15:24:00.000-08:00

In the loft
The solar panels are connected to an inverter (right - click photos to enlarge) , which converts the DC generated current into AC, using the "only clever bit" of the entire installation, according to the installer. This is mounted just inside the entrance to the loft. The interesting thing about the blue box is the "knock interface"... to switch on the unit and light up the screen you have to knock on the grey area. The screen tells you how many watts the panels are generating, and how much have been produced during the day, and how much in total.

In the kitchen
Downstairs, next to the consumer unit is a fantastic red knob, which isolates the panels. Above it is a meter which also shows how much electricity has been generated by the panels in total. During the first day of operation, the unit generated a total of 700W - disappointing considering it was bright sunshine outside all day, and the unit is spec'd at 1000W per hour. The installation engineer reassured me that as the sun got higher during the summer months the unit would generate much more - we shall see! During the second day of operation, when it was cloudy, the unit generated a total of just 200W during the day. Not good.

Metering
The electricity meter has been switched to an import / export meter. The meter gives the amount exported to the National Grid (ie total produced by the solar panel minus total consumed at the time in the house), and also the amount imported from the Grid. You can see the current cost clamp in the picture. This is interesting, because the current cost meter measures the flow of electricity regardless of which way it flows... so it reads 100W when the house consumes 100W, and it reads 100W when the house exports 100W. It was highly confusing turning on a light to see the amount on the meter decrease! So one of my Christmas projects is to connect another meter to the solar panel cable, and add the two readings together in order to get a sensible output... more on that later.

Solar Panel installation

2008-12-16T05:23:00.000-08:00

Here's a photo of the solar panels which were fitted to our roof today by Rayotec. They should provide us with 1kW electricity per hour in peak sunshine, but just around 100W per hour in horrible foggy weather like today. Tomorrow morning they will be grid connected, so any extra electricity we don't use will go back into the grid. With Southern Electricity giving us slightly more per unit than we pay for our electricity, there's still an incentive to use as little as we can. We expect the unit to just about cover our usage, over a yearly period. It cost around the same as having all our doors and windows replaced a few years ago and we're promised a similar length of service. So, we're just looking forward to the grand switching-on ceremony tomorrow morning!

Mind reading headsets

2008-11-24T09:04:00.000-08:00

So far, the mind reading headset market aimed at the regular consumer is quite small... I've come across the well advertised Emotiv headset, to be released early next year, and OCZ's "Neural Impulse Activator", shown right.

Unlike the Emotiv headset, with 16 different sensors placed on different parts of the head, this unit only has 3, all placed across the forehead, inevitably giving slightly less function.

In the first several hours of playing with the device, I managed to work out what successful calibration "looked like": - when the muscle signal graph didn't vary away from the "desired baseline". Sometimes this was very easy to achieve, and other times, just didn't happen. I discovered it worked better on one side of the lounge at home, than on the other. I also concluded that wetting the 3 sensors with water helped to a degree. Unfortunately the device needs to be calibrated every time it is used, which makes demonstrations rather hit-and-miss.

My first successful experience came shortly after calibrating the device. I was able to play the game of pong, using my eyebrows. By wiggling them fractionally upwards, the bat went up, and wiggling them downwards, the bat went down. And whilst I didn't ever manage to win a whole game, I did hit the ball enough times to win a few points, so it was not all chance.

My other successful experience was linking the muscle movements to keypresses, so I could use my eyebrows to navigate around Google maps. An eyebrow up-wiggle mapped to Page Up, and sent the map flying in one direction, a down-wiggle mapped to Page Down, and sent it flying in the other. Unfortunately zooming in and out was a bit beyond me, as the '+' key was unmappable.

So what about the mind-reading? Well, unlike the Emotiv device, it is not possible to "train" the device to recognise when you are thinking of stroking a cat for example. As found in this review, it should be able to detect general mood - OCZ's VP Dr. Schuette, says he can think of an expletive to make his Alpha 2 increase enough to trigger a "jump" action in his computer game. However I suspect this takes many hours of practice!

So for now, I shall continue playing pong with my keyboard, and wait for a few weeks when I hope to be able to get my hands on an Emotiv headset!

How to communicate if you can barely move

2008-10-14T13:30:00.000-07:00

My wife works with people who have had head injuries or strokes, who cannot move from the neck down and sometimes cannot talk; some are mentally active, which must be even more frustrating. Even in this day of great technology, letters are drawn on paper, and letter by letter the person moves their head from one side to the other side to indicate if this is the next letter in the sentence. As you can imagine, this takes a very long time, and it cannot be used to initiate a conversation.

Recently, I've been playing with a very good piece of software aimed at eye tracking called Dasher. This allows letters, which move in from the right of the screen, to be selected one by one by moving the mouse up and down and gradually forming a sentance. It is also predictive, so if for example you have selected H and then e, llo would be close together. It runs on multiple operating systems.

Secondly, my colleage Dave CJ has created a fantastic python script to control the mouse pointer on a Ubuntu PC using the accelerometer on the Nokia N95 via bluetooth.

So... by combining the two, I can tilt the N95 from side to side, to move the mouse on the PC and select letters in a sentence. OK, so balancing a phone on your head to do this looks slightly stupid, but if it helps open up a means of communication without any specialist equipment, or the help of someone else, it surely must be a good thing!

Back from Norway

2008-10-09T06:30:00.000-07:00

The day before our two-week holiday, we decided to go to Norway. We flew to Oslo (Torp) - and realised that the Torp bit means a 2 hour coach journey into Oslo centre, from the airport. We had a look at the technology museum - interesting to see the exhibits, though it would have been even better had more of the descriptions been translated to English.

We caught the (7 hour) scenic railway, reminding me of Dale as we passed Dale - to Bergen - a city which lives up to the description they give it - "the big city with a small town feel". It has a very nice harbour, and interesting shops. Eating out in Norway was just under twice the price of the UK - one night we managed a pizza, spaghetti bolognese and 2 lemonades for around 500 NOK, or just under £50.

The Flam Railway is well worth a visit which is, as they claim - "one of the most beautiful railway lines in the world". The end of September is the end of the tourist season for most things, but did not affect our walking much. In fact it helped, as we were able to extend our stay in this fantastic wooden cabin in Flam. It came with wood burning stove, and it's own rowing boat, and had a great view of the Sognefjord.

Our last couple of days were spent walking in Finse, the highest point on the railway between Oslo and Bergen, at around 1200 metres.

In summary, if you've saved up first, it's a fantastic place to go for big mountains, nice views, and good walking!

I've changed my name, thanks to Google

2008-09-15T06:37:00.000-07:00

Having a common name like Kevin Brown is a bit of a problem when it comes to getting a consistent, unique id on the various different web sites like blogger, twitter and the like. I have experimented with a range of ids... but over this weekend I decided I'd go for a more radical option - to change my name.

I had a search on Google and discovered that kevinxbrown, kevinybrown and kevinzbrown all returned no search results - which surprised me a lot! Have no other Kevin Brown's in the world got a middle name starting with these letters? So I've now adopted the middle name Xavier (well, what other names start with X?) Sorry to my twitter community who I totally confused over the weekend with my little identity crisis - ignore my invitations - I think I'm back in one piece as kevinxbrown.

Kevin Xavier Brown. :-)

Playing with Python

2008-09-11T13:52:00.000-07:00

Just for my own amusement, I've been experimenting with Python on the N95, and been amazed how easy it is - for example to make the N95 speak, you include import audio at the top of the code, and then write audio.say("hello") in your code. Taking a photo is nearly as easy, and using the accelerometer is also fairly straightforward. There are already many blogs and videos about this on the net.

My first very simple application isn't useful at all, but detects when someone picks up my phone and says to them in a kindly electronic way "Please do not steal this phone".

Emerging HCI technologies

2008-09-03T07:31:00.000-07:00

I find it very old fashioned and frustrating to sit down in the office every day and look into a box, whilst pressing my fingers rapidly on little pieces of plastic representing letters, and moving a small plastic box from side to side. It's not natural and can't be that healthy - ignoring the mandatory pizzas and Hobnobs consumed whilst doing so. The QWERTY keyboard was patented in 1878, and the mouse is 40 years old - so given the advancement everywhere else in technology, it's about time we got something else that beats it.

The good news is there are many exciting technologies which are either newly with us, just around the corner or have been reduced to a price which is accessible for the average user. I've been following some of them, such as embedded ViaVoice for talking to machines in a natural way, "mind reading" headsets such as the one from Emotiv aimed at gamers and due for release in December, wearable displays which allow you to augment reality with your own data - and mobile phone projectors all to be released in the next few months - as well as haptic user interfaces like this ultrasound device written about on the BBC today. I think the next few months and years will be very exciting! Just been playing with an OCZ NIA but that's a post all of it's own....

Image: http://www.flickr.com/photos/truebluetitan/273269366/

Why start a blog?

2008-09-02T02:31:00.000-07:00

Having been on Twitter for a while now, I've discovered that sharing some of the things I've been up to with the wider community has been very beneficial. I've made contact with lots of clever people, and it's interesting to keep up with their Twitters. I've blogged on my employers internal blog from time to time about the joys of Dojo, but blogging specifically about an area of work is all very well, until the line of work changes - which it does very frequently. Therefore I've started this blog, which I think will cover a wider range of interests. But writing this blog is only half of it - I'd like to read your blog too, so reply with a link, and I'll take a look!