Girton Labs Cambridge UK - Sensing Computers

Is this where the Kinect really started?

2012-01-11T02:26:00.000-08:00

Posted on April 15, 2011

This article was first scheduled to appear in The Independent in 26 February 2002.

Hmm, ultrasonic detection… navigation… that sounds like…

Amazing what you discover when you’re looking for completely other things. I was doing a search for “tablet” and hit this.

BY CHARLES ARTHUR
Technology Editor The Guardian

After the hands-free phone – the hands-free computer. Microsoft has secretly developed a prototype handheld computer that lets users navigate around large documents or images on screens by moving their hands or heads – but without touching the computer.
The system, developed at its British research centre in Cambridge, was last week taken by its British-born inventor to the US to show off at the company’s headquarters, to assess its commercial potential.
Rather than the usual mouse or keyboard, the new system – dubbed “Sonarscreen” – senses the movements by using ultrasonic emitters and sensors attached to the screen. While the sound is inaudible to humans, the computer can sense it and assess changes in the echoes as movement, much as a bat senses how near walls and objects are.
Its developer, Lyndsay Williams, hit on the idea when she was trying to navigate her way through a map on Cambridge on the Web to find a specific location.
The result – familiar to many Web users – was a frustrating episode in which she kept clicking to zoom in and then out and then left and then right, but kept missing the image she wanted.
“I said to myself, there has to be an easier way to handle large images and documents without depending on a mouse and without so many clicks,” she noted in an internal Microsoft briefing document that has been seen by The Independent.
Although Microsoft is best known as a software company, it has recently moved into making hardware with its Xbox games console, to be released in Europe in March, and says “major hardware manufacturers” will offer pen-operated tablet-shaped PCs, made to its specification, “in the second half of 2002″. The Tablets would be the ideal format for the new ultrasonic system – although some observers have cast doubts on whether the shape will prove popular with users.
Ms Williams, who has a long pedigree in computing – including having designed the hardware for the first sound card in an IBM-compatible PC in 1987 – rigged up a working model of her system in just four weeks.
Presently it is just a demonstration model, believed to be based on the Tablet PC, which is essentially a large screen with a computer processor attached. In one software setup, for map viewing, moving the face nearer to the screen enlarges the onscreen image, and gesturing left or right with the hands scrolls the image in the same direction.
Another setup will enlarge onscreen text as the face gets further away – so that one can view the screen, say for reading an electronic documents, from a comfortable distance.
“We proved the technology works and now we really need to test users to find out if this method of manipulating information is easy and comfortable for customers to use on a handheld device,” Ms Williams notes in the internal document.
But Andy Brown, research manager of mobile computing at IDC, a computing market research and analysis company, said: “It sounds interesting, but over the years there have been a lot of things that are invented, but very few actually materialise. The trouble is that their invention is driven by engineers rather than end-users. And though tests might find that people say they like it, if it’s at all expensive to implement then it won’t get taken up.”
However he said that it could have interesting applications once “pervasive” computing – where computers are built into homes and react with people constantly – becomes commonplace. “But that’s about five to ten years away,” he added.
Microsoft Cambridge estimates that the ultrasonic sensors cost no more than a few pounds each, and are just 3 millimetres across – which ought to mean that they pose no significant cost to PC manufacturers.
The Microsoft Cambridge laboratories were established in June 1997.

reinventing the wheel

2012-01-08T02:18:00.000-08:00

Biomimetic robot

The goal was to make a robotic device with the following goals.

very small and compact - 2" diameter
no external moving parts
very low weight - currently 22 g with battery and debug connector
can be made waterproof -wrap in silicone rubber and use inductive charging
will work on slippery surfaces
long battery life
collision avoidance using infrared
can be radio controlled or GPS driven

Here is a Video of a very early prototype made with an Arduino LilyPad.
It is based on using the movement of centrifugal printed pancake motors. Further work needed on accurate steering etc.
Prototype currently runs for > 2 hours on one recharge of 3.7V 150mAhr model helicopter Lithium Polymer battery.

End Jan, the graphene will be delivered to Girton Labs and so the low friction surface can be added onto the actuators.

So far applications are not much more than a cat toy but possible military applications.

Further reference - sidewinder snake movement

E Ribbon - Connect your body to the internet

2011-12-04T11:13:00.000-08:00

E-Ribbon, designed by Girton Labs, is a flexible silicone rubber ribbon, soft, safe, lightweight, wearable, washable sensing and actuating system for the human body. Applications include military for monitoring soldiers, sports, dance, medical applications (e.g. our Early Onset Epilepsy detection system using motion sensors), gaming, fashion, relaxation and massage. E-Ribbon uses individually addressable sensors such as accelerometers or temperature. E-Ribbon also provides waterproof touch switches, LEDS, and vibration actuators for pleasurable skin stimulation . The system can connect to the web via Wifi.

We have spent some time designing a waterproof and washable system, and it is based on transparent silicone rubber, working to IP68. There is no ingress of dust and there is 1 metre of water submerged protection. The E-Ribbon can be cut to length for attaching to textiles. We are integrating a novel non contact heart rate sensor can also be fitted into the ribbon. The high speed digital bus will also allow cameras to be added.

We have a working sample, off the shelf Red Green Blue waterproof LED lighting strip is the basis of design, but we can fabricate our own with custom flex electronics.

The flexible transparent silicone provides the IP68 enclosure and strip for washing in temperatures of up to 70 Centigrade .

You will see the current copper flex pcb has three tracks, power, ground and signal. The signal track allows multiplexed sensors and actuators with in future built in redundancy (multiple power tracks) for reliability. We plan to use a 2 or 3 wire flexible system with ground and combined power, data. This will allow approx 128 accelerometers on a strip (Red Green Blue LEDS shown in image) I2C is the serial protocol for sensors, but we can use a 64 bit addressable system for some sensors. Other actuators can be miniature pancake motor vibrators as used in cell phones, each individually addressable, The vibrator motors will be used for a safe and therapeutic relaxation system we are designing. Ripples of vibration can be applied to the body. The vibrators can also be used for feedback and communications for military personnel. Colour and temperature sensors can be used for a wound bandage sensor.

Typical I/O controller for flex circuit

E-Ribbon options for mixed sensors/actuators:

accelerometers - motion, fall detection
heart rate health
temperature sensing
waterproof switches
RGB colour sensing - wound detection
vibration actuators
sounders
Multi colour LEDS
programmable electronic fragrance - none waterproof strip

Advantages over current e-textiles for carrying power and data around the human body:

Can be retro fitted to existing garments and cut to length
waterproof
multiple sensors/actuator built in
washable

A Programmable System on a Chip (Cypress) will provide the main system Control.

Other options include connection of the sensor system to the web via Wifi or mobile phone via Bluetooth. Also using the silicone ribbon will allow easy integration into e textiles.

Video here of wrap around Red Green Blue LEDs underwater.

We also now have access to a heart rate sensor that can be attached to the body but works through clothing.

Early plans for this are a therapeutic relation system that measures heart rate and applies relaxing rippling mechanical vibrations to the person on their neck and back via a waistcoat design.

LED dress from Nancy Tilbury for Philips video here

References:

Intelligent Textiles Centre for Defence Enterprise

strip LEDS

Controller

PWM LED Controller Data Sheet (output only)

Example one wire temperature sensor

Arduino code example

Here are details of the 64 bit addressable sensor/actuator system from Maxim.

E-Ribbon designed by Lyndsay Williams, 4/12/2011.

More info sensecam@gmail.com.
Tel +44 (0) 7970 101578.

The Chocolate Chip - World's First Electronic Fragrance Semiconductor

2011-11-20T06:00:00.000-08:00

Wake up and smell the chocolate ....

Here is Girton Labs latest development of Programmable Electronic Fragrance, PEF designed by Lyndsay Williams. These are a development of our SenseBooks which were cookery books with images and associated scents for herbs and spices, tested by Girton Labs in 2008, research financed by EEDA.

Girton Labs design philosophy is that engineering is a combination of science and art, and our latest creation is this elegant chocolate button. We decided to use synthetic chocolate fragrance for the first test - who doesn't like the smell of chocolate?

Chocolate also has therapeutic properties as described here in Nature. PEF is based on a Cypress Programmable System on a Chip. PEF is part of the ThrillChip system, an arousal system for the senses that uses touch actuators such as vibration, heat, cold and nerve arousal, and now adding electronic fragrance , PSOC and PowerPSOC allows MOSFET actuators, 20 bit analogue to digital converter, (used for measuring biometric signals such as galvanic skin response and controlling dose ), programmable gain op amps, accelerometer interface, power management and microcontroller on one semiconductor. This PEF custom semiconductor will be available for licensing once prototype development is finished.
The PEF chip has actuators that can mix multiple synthetic fragrances in the cylindrical chamber as above. A 3D printed Lab on Chip is used as part of the well or reservoir design. A switchable matrix will allow up to 16 fragrances to be mixed in the chamber. If just 8 fragrances are used, this should theoretically provide a maximum combination of 40,320 fragrances, but needs testing.

So if there are 3 fragrance pots, A, B, and C there are the following 7 possible combinations.

A
B
C
A and B
A and C
B and C
A and B and C

The time on for each fragrance will allow more combinations.

Fragrance technology has been tried in the past for many decades but has suffered problems from "lingering" odours. This problem is controlled in part by the using of a living hinge on the plastic enclosure design with Flexinol control of hinge to open and close scent exit. The main method of stopping the lingering odour is via working and proven Girton Labs proprietary technology, Scent Reset (TM) , details available under NDA. Sensors such as accelerometers detect movement in the person and allowed a controlled fragrance output.
PEF is low cost and the cartridges are replaceable. The system is as thin as the battery, in this current prototype we use an Nano iPod battery, and so the technology can be used in mobile phones.
More details of PEF available under NDA.

A example of a chocolate fragrance is Angel perfume by Thierry Mugler - but this can never be switched off once applied to the body.

Angel perfume Thierry Mugler

PSOC PEF semiconductor

3D printed Lab On Chip

PSOC, image from Cypress Semiconductors

PSOC prototype controller from Cypress, actual chip is approx 8 x 8 mm

2007 version of ThrillChip

Memories are made of disks - Sunday Times 11th Sept 2011

2011-09-11T07:09:00.000-07:00

Memories are made of disks

In the future every conversation, every emotion will be committed to our computers’ hard drives. But some people have already started

Simon Cox Published: 11 September 2011

Dr Cathal Gurrin has been lifelogging for five years and finds it 'compelling' (HO)

Imagine everything you do being tracked, photographed and recorded; every person you meet, every conversation you have, every place you visit and even the exact route that you take to get there. A complete digital archive of your life, updated minute by minute and accessible at any time. It’s the dystopian future George Orwell described in his novel 1984, where “every sound you made was overheard and… every movement scrutinised”. But this isn’t fiction. It’s happening now, and guess what: you’ll be the person carrying out the surveillance and storing it all on your own home computer. Welcome to the world of lifelogging.

Unless you’re a technology expert you’re unlikely to have come across a lifelogger. They’re mostly software engineers, or computer scientists such as Dr Cathal Gurrin, a lecturer at Dublin City University’s (DCU) School of Computing. “It becomes compelling,” says the soft-spoken academic, “and it’s a difficult thing to stop doing once you start.”

Gurrin’s journey began as a two-week experiment in 2006 using a SenseCam, the lifelogger’s key tool. This is a camera about the size of a cigarette packet, which is worn around the neck and can be set to take photographs when triggered by such things as changes in the light, ambient temperature or body heat, or be primed to take a snapshot, say, every 30 seconds. It has no viewfinder or display, but is fitted with a fish-eye lens to capture almost everything in the wearer’s view. Gurrin’s snapshots are stored on a server at the university. His original fortnight of photographing has turned into a five-year odyssey.

Dr Gurrin’s “eureka” moment came when he met his new girlfriend in 2006. She recalled bumping into him in a bank months before they had started dating — and there in his digital archive he found the picture of the moment he first saw her. Now he wears the camera from the minute he wakes until he hits the sack. He says he wants to create “a detailed digital diary of everything; a diary of you, for you”. So he also logs his exact locations and movements — whether he is driving, sitting, standing, walking — and notes the Bluetooth devices, normally mobile phones, that he encounters.

Since he began he has amassed a collection of 8m photos, 2m GPS recordings of his locations, and 20m accelerometer recordings of his movements. So what does he do with all of this information? Gurrin’s digital archive is more than a massive exercise in narcissism. He is a human guinea pig for the research team he runs at DCU, which is developing the first fully functional search engine for lifelogs. Human memory is, of course, automatic. “In our brain it just works,” explains Gurrin. “You just recall things, but it is difficult to understand how people will recall memories on a computer.” It is imperative that Dr Gurrin’s team’s search engine actually mimics the brain, so they have been conducting experiments with cognitive psychologists at the University of Leeds on the ways the brain processes memories. Although it is “really, really challenging work”, Gurrin is confident that the first search engine will be ready by the end of the year.

Historians will have a complete record of your life. Imagine if we’d had this kind of record of Einstein, or Shakespeare, or HitlerThe overall aim is to create a perfect digital memory. Should you wish to pass it on to them, your children and grandchildren will have a complete record of your life. So, too, will historians. Imagine if we’d had this kind of record of Einstein, or Shakespeare, or Hitler. Eventually your lifelog will be automatically uploaded and stored on your computer for posterity.

Privacy, of course, is a big issue. “This is a research project, so I can’t possibly record everything in my life,” says Gurrin. “I switch it off if I feel that it would be inappropriate, or if someone might object to what is being recorded. It is their choice, they have no statutory obligation to get involved.” He draws the line at recording certain events that take place in the lavatory or in the bedroom.

“The logging process records 98% of life events — the other 2% would simply be embarrassing.”

There is a parallel, he says, with the advent of camera phones. When they first came out, people suggested that we would photograph everything, but we don’t. Rules as to where it is or isn’t acceptable to take photos have developed, and people tend to abide by these.

What is more, covert use would be impractical. “Remember that the SenseCam is actually a very visible device when worn, and we never try to hide it. It captures only two to three photographs a minute, which are not high quality; they are memory aids, not the quality of photos you would print for an album.”

The idea of a search-able personal archive isn’t new. It was first suggested back in 1945 by Vannevar Bush, the director of the US Federal Office of Scientific Research. In an essay for Atlantic Monthly, he set out a vision of a machine called a Memex, in which “an individual stores all his books, records and communications”. This could be accessed quickly and “is an enlarged, intimate supplement to his memory’’. The Memex would be a desk where material could be automatically photographed and placed inside. Bush also suggested that to get this information into the Memex “the camera hound of the future wears on his forehead a lump a little larger than a walnut. It takes pictures 3mm square”. He predicted: “As the scientist of the future moves about the laboratory or field, every time he looks at something worthy of record, he trips the shutter and in it goes.” Bush’s machine was designed with scientists in mind, rather like an iPad put together by Jules Verne.

So who will be using the modern version when Gurrin and his team finally crack it? Pretty much all of us, if Gordon Bell is to be believed. Bell is the granddaddy, the guru of lifelogging. He started in 2001 and predicts that within a decade we will all be lifeloggers. His seminal book, Total Recall: How the E-Memory Revolution Will Change Everything, written with a colleague and fellow lifelogger, Jim Gemmell, declares: “Society at large is on an inexorable path to total-recall technology and it is going to transform the world around us… and change what it means to be human.” Bell can’t be written off as some crazy futurologist. Now in his seventies, he is still a principal researcher at Microsoft. His boss, Bill Gates, calls him “one of the industry’s most original thinkers”. He is a man in demand, with technology start-up companies lining up to get him on board. When he and Gemmell, a fellow Microsoft researcher, wrote Total Recall in 2009, they set out to map the extreme lifelogging project — a huge undertaking, for which Bell started recording and storing everything he was currently doing.

His SenseCam would photograph every place he visited, every person he met. He began keeping records of the food he ate. He would record telephone conversations; make copies of memos and notes. He even wore a heart monitor and GPS tracker, which he used to work out that his angina was brought on by eating even a tiny amount of fatty ice cream. By 2020 he thinks this personal recording will be as commonplace as emailing is now. “I don’t think people realise we are part of a revolution,” he says. Integral to this revolution is the creation of what he calls our “e-memory”. He has said that these e-memories will be “objective, dispassionate, prosaic and unforgivingly accurate” compared with our “subjective, patchy, emotion-tinged, ego-filtered, impressionistic and mutable” biological memories.

In neurological circles such claims are treated with scepticism and thinly disguised disdain. “They are missing the point,” says Susan Greenfield, professor of pharmacology at Oxford University. “You can’t isolate memories; they are interrelated and upgraded and changed every time you visit that memory.” Baroness Greenfield challenges the belief put forward by lifeloggers that autobiographical or episodic memory can be improved and extended by e-memories, “They confuse memories with facts,” she says. “Imbued with every episode are other interrelated episodes. You’d have to download your whole brain.”

Less sceptical scientists have started to explore how data are affecting the way we remember and recall events. In July the journal Science published research by psychologists into the effect Google’s search engine is having on memory. It found we are less likely to remember information that we know we can find on the internet.

But could lifelogging help us to hold on to memories when our brains begin to falter? Researchers are already exploring the potential of lifelogging devices in neurological illnesses. In 2007 Addenbrooke’s Hospital in Cambridge conducted a study to see if a patient’s severe memory loss could be improved by wearing a camera to record and recall interesting events in her life compared with keeping a written diary. The results were encouraging. After a month the woman could recall 80% of the events she had captured on her camera but none of the events she had recorded in her diary. It led the researchers to conclude that the camera, “in addition to aiding autobiographical memory, has a number of other benefits for brain-injured patients”, and that it could help to answer such questions as, “How quickly do we forget?” and “How does our memory change with age?”

If governments got their hands on lifelogging data they would have a means to 'document every material event of an individual’s life'Lifelogging evangelists such as Jim Gemmell fervently believe the recording of all personal data will unlock some of the big secrets about our health. “That’s what excites me the most,” he tells me when I reach him on his holiday in Hawaii. He foresees a near future where our lifelogs will monitor our bodies through a multitude of sensors implanted inside us. These will continuously record data such as blood pressure, cholesterol and a host of other vital signs, feeding them back to our personal digital archive. These monitors will alert us or our doctors when they detect problems.

Of course, says Gemmell, lifelogging can be embarrassing if you share things. “A few weeks ago I had a friend prodding me about the potential negative impact on privacy of lifelogging. Then I looked on Facebook and found that she shared vastly more of her life than I ever do.

"She already had embarrassing things shared on the web, and I don’t. How ironic! So, for me, the way lifelogging might be embarrassing is if I appear in someone else’s lifelog and they share things I wish they wouldn’t. Gordon has logged the most about me, but he is even more private than I am.”

One lifelogger, says Gemmell, was embarrassed by the intrusion of the process. “He was sharing everything with his lab — note, the sharing that we would never do or advocate — and was embarrassed by shots taken in the men’s room. I asked him why. I’ve accidentally taken shots in men’s rooms, and you don’t see anything revealing. He explained that, although nothing X-rated is ever seen, he is embarrassed by shots taken from his point of view, seated — of his legs with his pants down.”

Being caught with your pants down could be the least of it. There is undoubtedly a potential dark side to lifelogging. This was explored by Martin Dodge and Rob Kitchin, geographers specialising in digital technologies, in a 2007 paper for the Centre for Advanced Spatial Analysis at University College London. If governments got their hands on lifelogging data, they warned, they would have “the potential for invasive profiling”, a means to “document every action, every conversation and every material event of an individual’s life”.

As lifelogs would record not just what you’re doing but who you’re with and where you are, they could identify and penalise “indiscretions, perversions and minor infractions of the law”. Dodge and Kitchin concluded that giving the state access to our lifelogs should be avoided at all costs, as it would encourage “an ultra-conservative society”.

This could be a selling point to large swaths of the community. Lyndsay Williams, the lifelogger and software engineer who invented the SenseCam, says only that what she calls “people without integrity” should fear scrutiny of their personal data.

If you have nothing to hide, you have nothing to fear? This same old argument was used to defend ID cards and tougher counterterrorism measures, but we weren’t convinced that only the guilty need to fear surveillance. To provide the reassurance that personal data archives couldn’t be used for profiling, said Dodge and Kitchin, lifelogs would need to be able to forget.

They came up with the idea of “an ethics of forgetting” for lifelogging, which would mean, for example, designing software that would allow exact details of a journey to fade over time, just as it does in our memory. Gordon Bell calls the idea “nuts”. As with so many of the tough questions about the consequences of lifelogging, he believes “society will need to come up with a suitable answer”.

We may be able to stop the state getting its hands on our personal digital archives, but it will be up to us to decide on this great drawback of lifelogging, the further blurring of the boundary between the private and the public. Lyndsay Williams says we shouldn’t be worried, as “we have no privacy any more”. It was something she learnt during her years at Microsoft. At the same time, she is aware of how intrusive lifelogging devices can be. “The GPS can track not just which house you’re in but which room.” That could be bad news for the millions who cheat on their partners. Imagine that your partner got hold of your lifelog and saw that the innocent trip to a neighbour’s house consisted of an hour in the bedroom.

And it is not just “people without integrity” who have cause for concern. In the course of his project, recording all Bluetooth devices near him, Cathal Gurrin found that in hotels he would pick up Bluetooth signals from other phones being beamed through walls from adjacent rooms, or from floors above and below him. When his students looked through the data, it appeared to them that these devices were in the same room as him. Try explaining that one to a suspicious partner.

Conversations have the power to come back and haunt us far more than pictures do. If the two are combined they could be lethalDr Gurrin is optimistic that the next version of Bluetooth will rectify this particular problem. What technology can’t do is to stop people recording us when we don’t want them to. That will be up to us to sort out. As he says, rules are already developing on what is unacceptable to record for your lifelog. For him it was gathering audio. When he tried doing it, even in experiments, he found “people really didn’t like it and basically they stopped talking to you”. Pictures were fine but he found that people had concerns that audio “will be stored, archived, found again and used against them in some way”. A group of colleagues chatting together might make a boring photograph — but their conversation itself could be more interesting. Suppose they are actually bitching about the head of the department.

Conversations have the power to come back and haunt us far more than pictures do. If the two are combined they could be lethal.

Gordon Bell, too, found that videoing someone or recording their voice could unnerve them, so he came up with a solution. “They didn’t know,” he confesses with a chuckle. “I have a file called ‘conversations’ but I didn’t use it in any way.” But if others follow this example and record surreptitiously, as Bell believes many will, won’t it create fear and paranoia? A version of the 18th-century philosopher Jeremy Bentham’s famous Panopticon (“all-seeing”) prison where the inmates would never see the “inspector” conducting surveillance, and could never tell whether they were being watched or not? Lyndsay Williams says with asperity: “You’ll just have to learn how to behave.”

Will this bother the ultimate target market for lifelogging: children and teenagers? I doubt it. After all, teenagers increasingly share most of their lives online in what Susan Greenfield calls “the solipsistic need to interrelate all the time with others”.

The technology being developed — mini video cameras, sensors for constant monitoring of your health — will mean teenagers will be able to share far more private and intimate details about their lives than the posting on Facebook of drunken photos in nightclubs or sunset holiday snaps.

Data about your whole life could be collected and stored online. This will be a valuable commodity, and companies are starting to realise the potential of personal digital archives. The food company Danone recently used lifelogging for research into demand for some of their drinks brands. And media companies are thought to have used it to track how audiences behave at weekends.

Cathal Gurrin has noticed recently that big multinational companies have “suddenly pricked up their ears and thought, ‘Hey, there’s something happening here. We’d like to get our hands on this.’” He is regularly meeting with big technology firms — although he won’t divulge which ones — who are trying to get into lifelogging.

Gordon Bell and Jim Gemmell are firm in their conviction that by 2020 lifelogging will have entered the mainstream. Cathal Gurrin says it “can’t be stopped”, for the simple reason that “the technology is already out there”.

So what kind of society will this encourage? Gordon Bell calls it “Little Brother”. A democratised surveillance, carried out not by one omnipresent authority but by millions of individuals. Let’s hope this isn’t the case. Unless, of course, you really want the future to be a version of the Big Brother House broadcast directly onto Facebook.

Freedom of Information Act - revealing £635 Million pounds of UK Tax payers money

2011-05-07T02:03:00.000-07:00

Released May 5th 2011 the spending of the Technology Strategy Board on UK Research and Development

News article in Research Fortnightly 18/5/2011

Spreadsheet here in Alphabetical order indicating £635 Million spend on Research and Development in 2500
companies. The TSB will write to all these companies to inform them their grant information is now public.

Rolls Royce received the largest share as leader of partners of £91Million
Let me know if any errors sensecam@gmail.com

Here is the letter about the Freedom of Information Request

Salaries for TSB staff are here.

Girton Labs Ltd were awarded a share (approx £67,000) of a TSB grant of £683,536 for SenseBulb technology but we declined it in Sept 2011. The SenseBulb is a novel heat sensing camera that tracks people and objects and helps senior people feel safe in their home. Text messages are sent when a person falls or leaves the kitchen with hob unattended. It is a working prototype as image on iPhone shows. The initial design of it's predecessor SenseCam was designed at Microsoft Research Cambridge.

more details here

The project on hold at present until funding obtained but Girton Labs is working on other innovative medical research projects at present.

I then decided to use the Freedom of Information Act to see which projects were successful using tax payer's money. You can view the 2500 companies in the spreadsheet and check yourself.

My thanks to the below for helping with getting the information and offering business guidance:

Andrew Lansley, UK Health Secretary and his staff
Ravinder Singh, Iain Gray , Richard Harrison, Graham Hutchins, Bob Reid, of the Technology Strategy Board
Emma Butterfield of Trading Standards, Cambridgeshire
PayPal Fraud Detection, UK and Luxenbourg
Colin Holloway, Clive Huntley, Sensory Systems, Swindon
Stephen Telemetry Associates Limited
Dr Nigel Harris Director BIME
Cambridgeshire Police
East of England Development Agency
Nominet
Others not named here, including Cambridge lawyers

It has taken 9 months since Sept 2010 for me to get this information and I thank my friends for their support.

Lyndsay Williams FRSA
www.sensecam.co.uk

Managing Director Girton Labs Ltd, Cambridge, UK
www.girtonlabs.com
sensecam@gmail.com
tel +44 (0) 7970 101578.

Cambridge Student Marches 24th Nov 2010

2010-11-24T05:26:00.000-08:00

Cambridge Student Marches 24th Nov 2010 in Pembroke Street, approx 1245 GMT city centre

more here:

and exclusive HD video here

video copyright Lyndsay Williams

TinySense - Micro GPS tracking real 3D objects on LCD screen

2010-07-01T00:05:00.000-07:00

Girton Labs have developed a technology to track to within 5mm the position of multiple physical objects on an LCD screen. The objects are placed on an LCD e.g. iPad (some testing on iPhone 4 ) with a graphic background. An example are chess pieces which the user moves but computer needs to keep track of. Other applications include games on pcs that use pieces e.g. Scrabble, models soldiers, ships. This sensing feature allows scores to be calculated automatically.There are music and audio applications with real control knobs on screen.
The technology is low cost, and works on many tested lcd screens. The system uses object recognition.
Similar devices for adding object tracking to lcd screens are Microsoft Research's Surface as here, but Microsoft's coffee table is large and expensive, (weight 150 pounds, cost $12,500) an iPad or standard non touch laptop is more suitable for consumer use in the home. iPad screen is robust for objects to be placed on with no fingerprints ( tested this on iPhone 4).
TinySense is like GPS Satellite Navigation but for the home. It is portable and handheld.

iPhone 4 with test control.
Here is a video of a static sensor in test.

Advantages over prior work are:
1. works on LCD screen with graphics background
2. very low cost, approx $100 per system.
3. resolution to 5mm or better (theoretical)
4. Each object has unique identity.
5. Very small objects can be used e.g. a MonoPoly games piece , with no batteries required for pieces. Battery power used for sensor receiver.
6. Multiple objects can be tracked.
7. The system can be scaled to very large displays and multiple sensors.
8. Sensing objects such as rotary controls and switches can be used.
9. Customer can fit the sensor to their own objects if they choose.

The system can also be used with printed electronics on paper (no LCD) and we are working with a Cambridge company on this system. An example is a MonoPoly board.

This research and invention by Lyndsay Williams was partially funded by the East of England Development Agency. It is a research project and not a production item yet. We are looking for developers and finance to complete.

This is the latest development of Sensesurface from Girton Labs.

TinySense and associated Intellectual Property belongs to Lyndsay Williams and Girton Labs Ltd.

For more details contact Lyndsay Williams on +44 (0) 7970 101578, or sensecam@gmail.com
Current project name: TinySense

Lorry explosion and fire in Girton Cambridge 22/6/2010

2010-06-22T01:31:00.000-07:00

First images taken 05:38 facing East Weavers Field garden view

We were woken up at 5am with 3 very loud explosions. All I could see from house was black smoke behind the wood. There was a eerie silence, no traffic, no birdsong, no sirens, and no commotion. It was very scary not knowing what was behind the wood. I called 999 and then went to investigate. The incident was 200 yards from home. These images taken on an A14 cattle bridge Girton, Cambridge at approx 05:30 facing East. A14 closed in both directions Girton. Street image taken behind hedge from our garden. There is some fire debris in our garden. There were no injuries, Cambridge Evening News has news on incident here.
Update Wed 23/6/10 here

Cleaned up facing West 10am

Incident 500 metres East of J31

View Larger Map

SenseCam human memory enhancement on BBC2 TV Eyewitness

2010-05-13T00:07:00.000-07:00

12th May 2010 SenseCam, human memory enhancement camera invented by Lyndsay Williams, for Microsoft Research on BBC2 TV as here The video starts 17 mins into program. BBC iPlayer for UK viewers only, and expires on 19 May 2010, but here is an shorter 8 minutes edited version here on Youtube

The program shows MRI brain scan of memory enhanced brain. Research by Professor Martin Conway of Leeds University.

I originally invented SenseCam for Alzheimer's as described in patent here

Image below shows memory enhancement in brain using Sensecam., click to enlarge.

Martin Conway and patient.

Contact Lyndsay Williams at sensecam@gmail.com or +44 ( 0) 7970 101578 for more details on SenseCam and later versions.

More on Lyndsay Williams' research at Girton Labs.

Philips moving into smell research

2010-05-01T00:28:00.000-07:00

Girton Labs has been working with Dr Jenny Tillotson Sensory Design and Technology for several months on computer control of smell emission and we now will be working with Dr Jenny Tillotson and Philips on new smell research as here, scroll down to Philips. This is an Arts and Humanities Research Council Knowledge Transfer Fellowship .
Summary of project : Smell The Colour Of The Rainbow: Sensory Systems For Wearable Product Technologies To Reduce Stress And Sleep Disorders here.

From http://my.arts.ac.uk/e_brief/archive/30april.htm

Bright sparks

Jenny Tillotson, Senior Research Fellow at CSM has been awarded an AHRC Knowledge Transfer Fellowship (£249,000) to work with Philips Research in Cambridge supported by Dr Raymond Oliver, Senior Research Fellow in Active & Interactive Materials from the RCA. Over the course of a year, they will engage in an active KT programme to explore opportunities resulting from her academic work in areas of interest to Philips Research. The project will also involve Lyndsay Williams' of Girton Labs (inventor of iPhone sensors and Microsoft's ‘Sensecam’ for Alzheimer’s) and smell expert Professor Tim Jacob from the School of Biosciences, Cardiff University.

The iPhone Achilles' heel

2010-04-30T11:03:00.000-07:00

There has been enough comment from others re the loss of Apple's iPhone 4G on a bar stool and ending up in Gizmodo. People haven't asked, why was it found on a bar stool? Would person have placed it there and just walked off? Unlikely. It could well have slipped out of the pocket. When I last visited Apple in Cupertino re sale of patent of my Invisible Phone for iPhone sensors, the Apple person I met with was startled that my iPhone was in a pink silicon gel case. I put it in this case as iPhone as slippery as a bar of soap. ( I dropped it first day I had it on concrete, while trying to take a picture.) Apple attorney said the designers would not like this as spoiled the smooth surface of iPhone which is designed to slip in (and out) of man's back pocket of trousers. I replied that ladies do not normally put phone in back pocket and I said if I did, I would not want it to slip out. Maybe this is what happened with recent loss of prototype.

There are far easier ways to find out what might be in future Apple products than checking out stools. A perusal of my 1997 BT SmartQuill patent Apple bought recently covered here in The Guardian and patent as here shows a portfolio of sensing systems for phones. This is all in the public domain.

Interesting to note, that one of only two working SmartQuill phones as above was stolen from The Millennium
Dome in 2000.
I am now designing sensors for phones for the next 5 years, including 3D hand trackers, not for Apple but a well known Far East phone company and others if interested . I am not allowed even to say if a Non Disclosure Agreement exists. I also have a system that will alert person if more than a few feet phone their phone. This is old technology based on radio systems that alert if a parent losses a child in a busy shopping centre.

I am still looking for the lanyard fitting on my iPhone....

SmartChill - No More Headaches

2010-04-30T02:07:00.000-07:00

Girton Labs researches and prototypes new methods of healthcare and pain management. One serendipitous discovery from the last device I invented before leaving Microsoft, ( ThrillChip as here, a feedback device for replicating fear in games controllers ) was to find new way of reducing the pain for some (not all) types of headaches or sports injuries. Some pain can be removed by a bag of frozen peas or instant ice gel coolers (£1 per usage) but not always to hand, e.g. when traveling and far from home. The SmartChill ( working project name) devices uses a Peltier heat exchanger to reduce the surface of the skin by approximately 15 degrees Centigrade. You can see the Peltier in image below, it is the square part. One side effect of a Peltier heat exchanger is when the power is switch off , the heat will flow back into the Peltier device, which might be close to the skin. A novel design of heat sink allows a safe method of chilling the skin and avoids this heating problem. Headaches take about 5 mins to be reduced when tested on a subject. The sensation of simulated ice on the skin is so relaxing. The device works off 2 x AA alkaline or rechargeable cells and is charged by USB or 12V car cigar socket. The device will allow about 1 hours operation before recharging. That covers several headaches and costs a lot less in recharging than buying and swallowing pharmaceuticals. It also avoids paying license fees to pharmaceutical companies.

Safety mechanisms in the built in computer hardware and software ensure the heat "sucked out" of skin is dispersed of safety.
Device is also very useful for sports injuries.
A built in recording devices logs the time and date of application of chiller, can download via USB to pc, so useful for research with working out what or who causes your headaches. This would work well with clinical applications of Sensecam.

Girton Labs needs a route to market for this, as customers want it. A retail cost of less than £50 is practical.

So what happened to the ThrillChip? It never got patented for intended use as a Xbox games controller, and after I left one of the UK leading games company wanted to use the technology. Microsoft choose not to patent it so could not license the IP.

For more information, contact Lyndsay Williams sensecam@gmail.com or tel +44 (0) 7970 101578.

New Hand Tracker for mobile phones

2010-03-10T22:59:00.000-08:00

One day in May 1997 I woke up with an idea for a new type of mobile phone, that had no conventional buttons, but it would be controlled by invisible sensors, i.e. accelerometers, screen touch sensors , face tracking devices. I approached Peter Cochrane, Head of BT Labs, Martlesham to see if they were interested in it and could they fund me to prototype it. They did and we patented in 1997. Patent also covered hand writing recording, cameras, GPS tagging and cloud computing. Microsoft liked this new type of phone and gave me position as a Reseacher at Microsoft Research Cambridge 1998-2007. Microsoft were offered this patent but didn't buy it.

Eventually, 11 years later this phone Patent sold to Apple in 2008 as covers technology used in the iPhone, here is some of the story

I left Microsoft in 2007 and founded Girton Labs Ltd, Cambridge. Now in 2010 we are inventing and building prototypes for mobile phone technology for the next 5-10 years. There is so much competition now with phones, a brand new type of novel hardware needs inventing to differential the rest from the Apple iPhone.
One thrilling new sensor design for mobile phones I can't talk about yet as confidential for phone company. But here is one I can. Here is a video of our new hand tracker. We have prototyped a proof of concept model:

If video doesn't fit on your screen, please click on link below:

http://www.youtube.com/watch?v=JrGXxKISOao

The video shows a hand being tracked with a 4x4 heat
array, of possible use to mobile phones and tablet pcs.
A Digital signal Processor decodes, left/right up/down hand movement .
Applications for this new User Input include web page navigation, hand
gesture control and power on/off for mobile phones and tablet computers.
Other applications include sterile medical controls and controlling simple
instruments in cars like GPS. It works with gloves unlike some hand trackers or touch screens.
The technology also includes these feature being done by a phone's built in camera
but the advantage over using ordinary built in camera to track hands, is very low
processor overhead, (uses 4Mhz micrcocontroller for Digital Signal
Processing in prototype) very low power and can work in low light
levels.
Cost of sensor approx $5 in large quantities, Intellectual Property
owned entirely by Girton Labs Ltd, Cambridge.
Please contact Lyndsay Williams, sensecam@gmail.com for more information.
www.girtonlabs.com

BusyBadge - Sensebadge for personal energy monitoring

2010-02-07T03:32:00.000-08:00

I have designed miniature computers for personal energy and activity monitoring for many decades. One below for Microsoft Research was an accelerometer, radio beacon and data logger, around 2001, SmartMoveX here. Research was with Microsoft Research John Krumm, Greg Smith, paper here.

This later got transformed into device with a camera, Sensecam as here. The University of Dublin have recently started some research with SenseCam on Carbon Footprints.

Last year I proposed a device for personal energy monitoring for medical uses that was very small, unobtrusive and could be used and worn 24/7 with no battery change etc. It is like a "well being" badge that display's personal activity in the day. The above sensing badge with paperclip ( real size model ) is a prototype, it has 3 axis accelerometer, data logger, 32Mbytes FLASH memory, radio transmitter, and tiny thermal camera 4x4 pixels and activity lights. It is a development of a SenseBulb I designed for Assisted Living but much smaller.
SenseBadge lasts about a week before battery change and upload of data to PC as a graph. As well as recording personal movement and exercise, the thermal camera can record ambient temperature, but also capture simple 4x4 pixel thermal image in front of person, i.e. record the time person in front of a stove, TV, eating, PC, human. These home appliances have a thermal signature that can be captured. Unlike normal digital thermometers, thermopiles captures very fast temperature changes, around 100 times/ second, see below 4x4 colour map.

As the data is time stamped it can be synchronised to camera images and GPS. As the thermal signatures are fairly simple they can be interpreted and classified using a computer, e.g. Excel, rather than complex image processing. Themopiles also avoid the problem of conventional temperature sensors worn on the body, these reflect the temperature of the users body (not ambient temperature) as wearable badge warms up , e.g. as on SenseCam.
Unlike passive data loggers, this one has optical feedback (LEDs changes colour) to provide rolling history of how much energy is consumed or wasted in the day. Every person you met that day will know at a glance if you have been a couch potato (or ill), sat behind a PC, eating, or "chronically busy" person ;-). You can walk into a meeting and people can glance at you for status update. As badge detects and records proximity of people, you don't need to say sorry, been busy "been in meetings all day" as display can show that at a glance. A download at end of day can classify percentage of time spent in each activity, e.g. time spent with other people. A sequence of spinning LEDs on badge speed up based on "busyness" measurement of last few hours. This can be useful in medical care of patients in a hospital to check if they are being cared for, e.g. accident and emergency , "lying flat on trolley for 2 hours and unattended". Executives would like it as faster spinning of LEDs might indicate "importance".

Some people say, well that can all be done with a mobile phone, but they are not fitted with simple thermal cameras (which can reveal more than optical cameras and accelerometers ), the badge is very low powered and can be worn at night. Badge also gives visual feedback of a rolling summary of the day's history of activity.

Image of UK Prime Minister who might need a BusyBadge.

Girton Labs Tablet Computer 2009

Badge Specification:
Sensors:
3 channel accelerometer
4x4 thermal camera -50C to 350C, sample rate 100Hz
digital light meter
UV sunlight detector
magnetic proximity sensor (optional)
Carbon Monoxide (optional)
Computer:
up to 32Mbyte FLASH memory
6V Lithium coin cells
serial port to pc
LED array for real time display of day's history of events, red, orange, green, blue, white, flash frequency indicates activity.
radio transmitter (optional) 433Mhz
FLASH Microcontroller

The practical side, Girton Labs has implemented all the components of this, i.e. hardware, software, algorithms, pcb, CAD case. The applications range from personal energy monitoring, medical applications, Assisted Living.
Contact Lyndsay Williams sensecam@gmail.com. or +44 (0) 7970 101578

What's new in touch sensing mobile phones?

2009-12-12T05:48:00.000-08:00

I have worked on a lot of touch sensing technology. I am particularly interested in the remote transmission of touch and feeling over the web. One device I invented and built was a "heartbeat mouse". This allowed your heartbeat to be measured and replicated as a physical beat or thump on a computer mouse with someone you are communicating to on the web, e.g. a friend. It was possibly useful for Instant Messenger type applications. I build a demo unit when I worked at Microsoft, the heartbeat being detected via optical sensors methods and the beat replicated using a magnetic mechanical actuator. Most people liked it, others thought it wacky, but I never patented it, and now in the public domain, so cannot be patented. It is research that can be shared. Microsoft has deleted the web page that described it so I can't show the original image.

However I am still asked about this design and how it might be useful for touch sensing applications.
A possible answer is via mobile phones, you are holding the mobile phone and talking to friend but can it replicate the feel of holding hands? The mobile phone is a substitute for holding someone's hand. You could choose to share your heartbeat and also that of remote friend by holding hands. The mechanical actuation technology is much more advanced now, the Microsoft version just used an modified oscillating relay with pulse width modulation (PWM) control to replicate feelings.( Curiously, the prototype was rather good at synthesising the feel of a cat purr , as the software control of the PWM was very fine. ) There are more applications for lie detection, medical apps, games and other fun ideas. It is practical to transmit and replicate skin temperature and conduction.
The only reason the project resurfaced is now the new "smart" actuation technology I have been testing for another job has been built, gone are old solenoids and other magnetic actuators. The new technology provide much more subtle vibration finger stroking effects, is silent and very low power.
Email me if interested sensecam@gmail.com
Lyndsay Williams

Open source sensing camera for Alzheimer's ?

2009-12-04T02:43:00.000-08:00

I designed the first SenseCam for Microsoft Research in 1999 and it was released to the public in approx 2005. I left MSFT in 2007 and since then I have received a high demand for sensecams from brain injury people. Microsoft only have a limited number of Sensecams for loan only to qualified people, e.g. PhDs and doctors, not ordinary people. They are not for sale. The good news is that Vicon are releasing the Revue which is similar to a 2005 version sensecam, but only in limited numbers at first for research. It concerns me to get long emails from carers who cannot get a SenseCam. My motivation for releasing a new sensecam is simply to help people with medical problems.

Girton Labs Ltd Cambridge is now a UK government accredited Research and Development Establishment for science and technology. We are considering releasing an open source version of a new sensing camera. This is due to the high demand we have for sensing cameras from brain injury people, so we have decided to possibly release an open source, free of charge design of a sensing camera. Girton Labs does not plan to profit from the sale or manufacture of this camera. Open source hardware will allow release of schematics, system diagrams, software routines and sources of the specialised components. Off the shelf software will allow rapid viewing of images.

This new camera uses different sensing methods from the Microsoft SenseCam patent. It will enable the design to be released into the public domain in the near future. We need to check with our lawyers first however with respect to patents for my design for Microsoft of the first SenseCam and subsequent patents.

The camera captures audio (as customers asked for this) , high resolution stills and video, uses 8G memory cards and is wearable. The video compression allows 1:100 compression of significant events in the day with capture of events that were hard or impossible with sensecam, one example being a dietary application. A novel system of sensors can capture a data record of food and drink eaten. This is currently on trial with a wearable device to record alcohol intake. The serendipidous discovery and testing of a multi sensor system to achieve this is delighting us.

Another very important improvement on the first sensecam is the ability to detect compliance with taking of precription medication.

We are considering releasing the design with Girton Labs getting no profit from the sales and manufacturer of the camera, simply to help people with Alzheimer's or other medical problems who have requested cameras.

I am very respectful of Microsoft Research Cambridge lawyers and their Intellectual Property and so the design uses very new sensing technology. Girton Labs is the first company in the world to use this new sensor device. It combines the functionality of the 4 sensecam sensors (acceleration, light, heat, temperature) into one tiny MEMs device, which costs around 3 Euro on the cost of a normal digital camera. The accelerometer is eliminated now, so results can be captured in moving vehicles with more accuracy.
A wireless link with a mobile phone allows GPS location tagging etc. The sensecam technology is not rocket science, it is a very simple system of sensors and camera to allow image compression and therefore easy for others to build with readily available parts.

Email Lyndsay Williams at SenseCam@gmail.com for more details.

Microsoft Research 2009 human body power patent application published again

2009-11-09T00:44:00.000-08:00

In 2004 one of my Microsoft patents was awarded for "Method and Apparatus for transmitting power and data using the human body"

Explained in this article in The Guardian 2004,
"Computerising the body: Microsoft wins patent to exploit network potential of skin" as here
In simple words our design (Williams, Vablais and Bathiche*) was to allow devices like headphones for your Ipod be attached to the ears without using wires or batteries in the earpiece. It is also useful for medical sensors to be attached to the body without wires. I heard after I left Microsoft that the US military were also very interested in using this technology for their soldiers. The prototype I built and tested on my daughter in 2000 allowed audible music tones to be transmitted to a loudspeaker on her skin when I touched her, no batteries or wires on her.

It is interesting that Microsoft's lawyers are still working on this technology 9 years after it was first prototyped as patent has additional wider claims published July 2009, example as below.

From the US Patent Office Patent application published July 2009

Claim 1 " A system that facilitates reduction of wiring in connection with networking portable devices, comprising:a plurality of devices that are electrically networked together via a human body, which serves as a transmission medium for at least a subset of the devices."

Will Microsoft license this technology?

*Steve Bathiche was one of the inventors of the Microsoft Surface computer.

Drawing above done by Microsoft's lawyers for the patent.

Prototype wireless sensing camera

2009-11-06T01:06:00.000-08:00

Here is Girton Labs's working prototype wireless sensing camera , transmits encrypted wireless images to a PC, lasts 1 year on small lithium cell, first application is for Assisted Living, i.e. those who live on their own. Captures and compresses all interesting daily images. Every room could have one. Should be reasonably to have it button sized, (one shown is a working proof of concept). Imagine it being like nerves all over your home, caring for you. Thermal images are scanned and compressed for interesting events in the home, people, doors opening and closing, fridge activity, cookers left on, radiators, fires, tv on/off. Useful also for energy management, can see heat from appliances switched on and wastage. Alternative version send text messages to cell phone on events. Could also send messages/images to Twitter. Other versions of camera can transmit normal wide angle images.
Current prototype sends colour thermal images to PC.

Frequently Asked Questions.
Q. How is this different from SenseCam that you designed for Microsoft?
A. It uses a totally different method of sensing, patent filed, and so is new Intellectual Property.
Q. How else is different from the first sensecam?
A. I designed SenseCam for people with very severe memory problems. A person with Alzheimer's is forgetful and might mislay the camera and all of the day's recorded memory, this camera transmits data to a central server or PC. As camera has radio transmitter it can also be located by it's signal if lost, e.g. hidden under a cushion. There is no limit to number of images as transmitted to a large server.
The camera can be worn or fixed on ceiling of room.
Q. How do you get such long battery life?
A. Innovation adaptive clock technology, based on how mammals conserve power. Details available on signing of NDA.
Q. When is it available ?
A. Currently due for delivery to first customer this year.

Designed and built by Lyndsay Williams
More info sensecam@gmail.com

Image shows test thermal map

SenseCam - a new type of video compression

2009-10-31T01:32:00.000-07:00

Some SenseCams I built

image on right shows SenseBulb messages

There have been quite a few articles re Sensecam in press recently re Microsoft licensing it. It is more than just a time lapse camera. I designed it initially in 1999 as a new type of video compression system to capture all the interesting moments in life - but leave out the repetition and dull bits. Many of us sit behind PCs all day and there are better ways of capturing the pc desktop see Microsoft's MyLifeBits as an example. (MyLifeBits does as a whole lot more, but this is one interesting feature of it, here is the book) Patent introduction above shows my 2004 invention of Sensecam for Alzheimer's.

In SenseCam, sensors such as heat sensors, to detect people, light sensors, to trigger a sequence of pictures of rooms as person walks through doorways. An accelerometer can detect user moving and capture image and also reduce blur in the image. A microphone is triggered on loud sounds and images captured. Tagging data with GPS as in my SmartQuill patent here. Even a heartrate increase can be logged with SenseCam images. These events may happen about once every 20 seconds and played back as a continuous sequence of images at about 5 frames per second. That gives an approximate video compression ratio of up to approx 1:100.

Here is a 2007 video example of my bike ride, the compression is around 1:15 here as quite a lot of action events when cycling. It is from Girton, Cambridge, up Huntingdon Road, past the Astronomy Labs to Microsoft Research.

A less dynamic sort of day will allow compression of around 1:100 or better. My "normal" days are probably a bit like many of us engineers , behind a PC, writing and debugging software, hardware debugging with oscilloscope and soldering iron. (I ignore email if I am debugging software as interrupts train of thought).

The SenseCam also captures around 400% more of the view as it's close to a fish eye lens view 104 degrees compared to around 50 degrees on normal camera.

JPEG comparison is a compression ratio of approx 1:100.

What next after SenseCam? After I left Microsoft I set up Girton Labs and designed SenseBulb technology to capture all that the Sensecam missed. It is a camera with heat imaging and captures life in all it's minutiae. It's first application is a camera that replaces a lightbulb and sends mobile phone text messages of activity in the house. Thermal images are replicated as "colour shadows" and so no real privacy issues. Mundane events such as back doors left open, fridges being opened, people moving around the home, eating, sleeping, TVs switched on indicate the "aliveness" of a house and it occupants. This is useful for Assisted Living, Alzheimer's and people living on their own. A family member will get a text message saying fridge door opened 3 times this hour, that means all is well.

I guess some people would find it useful to text or Twitter "I'm home" completely automatically via SenseBulb rather than manual typing. SenseBulb transmits encrypted text messages messages via GSM network so as not to inform burglars, Twitter is for all to see. I have prototyped Sensebulb V1 and its fun.

A bit more here on SenseBulb. (Patent filed) Page very out of date as current version V2 with colour pixel display now being tested.

Email me if any questions on this technology at sensecam@gmail.com or call me on +44 (0) 7970 101578 (Cambridge UK).

Girton Labs's new sensing camera

SenseCam Cosmetics

2009-10-30T13:47:00.001-07:00

The SenseCam case has had small changes since my first minimal design of 1999 then 2005, case molding by Microsoft Hardware Group.

Why so minimal? I design all my computers with no on/off switch, I use sensors like accelerometers or light to power on/off design.

From left to right, you can see the 104 degrees wide angle lens, enough to capture image of your shoes, LED status indicators, Passive Infra Red Sensor (PIR) and light sensor. The bump is for the AA cells, a lot easier to find in a hurry than a Lithium charger. I am interesting to note that the new Sensecam still uses a huge filter for the PIR.

The lanyard was nice and simple - until someone decided to "upgrade" it to magnetic catches using neodymium magnets. I had to quickly email one of our principle Microsoft Researcher users of sensecam, who had a heart pacemaker, to not be alarmed. Neo magnets are 5 times more powerful than normal magnets, and used to control the settings of pacemakers!

an early sensecam 1976

2009-10-23T02:18:00.000-07:00

I started experimenting with sensecams when I was 17. I had left school, and this was a paying vacation job between school and going to university. My other paying jobs were writing for Practical Electronics and Electronics Today International. The company was Baxall Electronics, Bramhall, Stockport, Cheshire, and consisted of me the first employee and Malcolm Baxter the boss, and owner Don Hall . The company was based in a small wooden shed. Malcolm had invented "Watchman" a Eurocard size board full of TTL and analogue components that could be used to set a zone on a video image to send alerts. Typically a control knob would be used to set a maker over a door in the view, and so video recordings of intruders could be captured. I did the testing of these and other video devices. Vidicon tubes keep burning up. I confess I never did tell Malcolm it was my error. I learned a lot about signal to noise ratio in analogue video circuits, great fun. Malcolm was ingenious, the video signal and power went down same bit of cable. My job was to optimise the transistor circuit for the best signal to noise ratio. Baxall Electronics grew to become one of the largest supplier of CCTV in the UK and the world, with several hundred employees but was taken over in 2007. Cameras were initially put in the London Underground. 33 years later the UK now has more cameras in public than any other country in the world. You might be viewed up to 300 times per day on London Cameras.

Roll on Microsoft's Sensecam.

Why no audio on the SenseCam?

2009-10-21T01:28:00.000-07:00

Many people ask why there is no audio on the Sensecam when I invented it for Microsoft?

It can be done, one, it is easy to use a standard digital audio recorder (e.g. maybe the new Apple ipod nano) to record audio with time synchronised information to sensecam video. Two, my concerns over privacy issues, conversation is likely to be inhibited if you know audio is being recorded. Some people just dry up and you never hear their best (or worst) jokes! Third, I did design an audio compression algorithm for Sensecam. This uses snippets of sound bytes from the day, e.g. train sound, London Underground, car engine sounds, laughter, footsteps, kettle sound etc to add an atmosphere of sound, and adventure to the video. It did not record conversation. The video stills are playing back at about 100 times faster than the days' events so the audio needs to keep up. I was developing this when I was deleted from Microsoft, and the work never got patented. It was based on detecting sounds over a preset amplitude and an FFT filter to pick out interesting sounds of artefacts. There was a microphone in early version of sensecams, see it in image, by the strap.

The iPod nano allows users to capture audio as here The Nano also captures video as well.

It is interesting to note synchronised music can be added to SenseCam videos, here are some.

First ever sensecam video here and another here

Disclaimer, I am a Microsoft Shareholder

I work part time for Apple Inc as Consulting expert

on my patent sold to them re iPhone sensor technology as here.

Lyndsay Williams

www.girtonlabs.com

sensecam@gmail.com

A brief history of Microsoft's Sensecam

2009-10-16T13:13:00.000-07:00

Oct 15th 2009, Microsoft finally license Sensecam see here. This is a sensing camera to help people with Alzheimer's.

(Image shows early version of SenseCam)

I invented and built the first sensecam in 1999, my patent filed 2004 as here , first paragraph explains my plans for helping Alzheimer's .

I had been approached by Roger Needham, then head of Microsoft Research Cambridge in 1998, to work for Microsoft Research after invention of my SmartQuill, a sensor telephone appeared on www.slashdot.org , this technology now used in iPhone. * See end of article.

Why did I invent SenseCam?

I wanted to help a person with memory problems who was always losing keys, also a dentist friend of mine, Dr Jane E. Jacott, Woodford, Cheshire, wanted some technology for time and motion studies of how dentists spend their time, sitting, standing, doing surgery, paperwork, etc. This really inspired me and concerned me for some months, as I wanted to help her. I also have lost a few months memory after being hit by a car . It is torture to lose your memory after a car crash. I think it was so much worse for Mum & Dad as they knew all but I lost my memories for about a year.

I built the first SenseCam using an Analog Devices accelerometer, a PIC microcontroller a digital camera that could take 64 images, a good capacity for 1999! It was hand soldered on Veroboard and programmed in PIC assembler by me in 4KB of memory. I attached it to front of my bike basket to capture images of all the drivers in Cambridge who kept hitting me. I curiously ended up with more pictures of ladies not looking after children in buggies and buggies ran in road and I got pics of wide eyed scared babies when I braked ! (accelerometer captured braking and triggered pic) But a lot of the images were poor and blurred and pic of a car bumper is dull. I searched for months for a suitable tiny fish eye lens and found one , this allowed a very wide angle (104 degrees) image. So wide angle it could even remember the shoes I wore on the day! I put the project on hold for a year or so until I found a camera module with 128Mbyte of memory. Meanwhile I was working with John Krumm, of Microsoft Research Redmond on people location badges using radio and accelerometers. This worked like indoor GPS. I have to say at Microsoft Research, my employer, there were many other supportive folk, but Andrew Blake was one of the the best. (Roger Needham also but sadly died,) Andrew knew I had being doing experiments with accelerometers on the body and prompted me to do a device with lots of raw sensor data output so the machine learning people could analyze it. Perhaps add a camera? The key bit to getting sensecam working was my algorithm for recording a sequence of rooms. I said to my then husband, who kept loosing his keys, which room where you in? Library, car, hall, lounge..retrace your steps. I designed software and hardware with a light and colour detector, accelerometer,temperature sensor and passive infrared detector. When we walk through a door, a step change in light occurs, i.e. a door frame detector, so an image can be triggered. Person in front of you, triggered by heat, captured image. Accelerometer, used to capture pic when you turned around, so pics all blurred! But I turned this bug around to capture image when person *not* moving so steady image. This was all done in an 8 bit PIC microcontroller, I did the hardware and software for this and James Srinivasen did the memory card software and other clever bits.

Trevor Taylor did a lot of the early pcbs and hardware. I designed and built the hardware and software. There were many late solitary nights with soldering iron and oscilloscope. Stuffing all my rat's nest of hardware into a case the size of a match box was hard! It was very embarrassing when early SenseCams burst into flames due to lithium battery recharging issues and the whole of Microsoft Research (then in offices in Petty Cury, Cambridge) was evacuated.

Here is first pic taken with sensecam,

Kings College, Cambridge

Grass is pink as I took out the Infra red lens, so the chlorophyl in grass looks red.

I had about two weeks to get a demo working for a Microsoft trade show, Techfest in Seattle and Bill Gates wanted to see it! I worked late weekends, luckily my husband looked after daughter. My managers did not want to show it at Seattle, so I had to be persuasive. Bill liked it, smiled a lot, and said he would like one for his pet dog to check how it was treated when it visited the neighbours. I hand built about 18 sensecams, then my boss said could I make another 100 including some for training American soldiers in Iraq as mentioned here? I said there are better people than me to do production engineering, suggested a company and that I didn't support the war in Iraq. (My boss was North American, I am British) At my termination meeting in 2007, this was mentioned as a reason for making me redundant. I left and I did next version of sensing camera, which I am testing today. Loopholes in my patent allow new inventions.

I must mention the time I was honoured by Gordon Bell visiting my office around 2000. Gordon said, never throw any emails away, this was a flashbulb memory moment for me, I was sat in my chair and Gordon stood at my office door. Gordon has continued to be of great support to me and my memory projects before and after I left Microsoft, thank you. Some of Gordon's and Jim Gemmell's work with SenseCam covered in this book Total Recall.

Also was very pleased that Narinder Kapur of Addenbrookes approached me to help with memory problems (wanted SmartQuill but I suggested SenseCam) and Emma Berry who did the neuro trials with patients with only a few hours of memory.

Microsoft Seattle helped me with later case designs.

Today, (16th Oct) I am happily testing my Sensebulb, a sensing lightbulb that sends text messages if people have kitchen accidents, fall over, leave the cooker on, really for Alzheimer's . Chris Curry, who set up Acorn Computers after parting with Clive Sinclair, approached me re SenseBulb, and I am happy to be behind an oscilloscope debugging my software.

Here are some of the my SenseCam videos click here

I will update this blog in case I have forgotten anyone, email me at sensecam@gmail.com .

Thanks also to Jack Schoefield of The Guardian for this article about my work.

Anyway, me and my team of trusty hardware/software engineers, many from Microsoft, past and present, are available at Girton Labs if you want help or questions with your latest projects. I am at sensecam@gmail.com, or cell +44 (0) 7970 101578.

* SmartQuill telephone patent with sensors? What happened to this invention of mine from 1997? Apple bought the patent in 2008 as a lot of the technology of this patent in the iPhone.

Lyndsay Williams

www.girtonlabs.com

Microsoft to license Sensecam

2009-10-15T02:58:00.000-07:00

I am very pleased indeed to see the SenseCam become a product, this is great news for customers. I invented the first Sensecam in 1999, wanting to help people with severe memory problems. It is very good news that Microsoft Research can license this technology.

Licensed to OMG, Oxford Metrics Group, who do the special effects for the Harry Potter films.

OMG shares went up 26% on announcement of this deal with Microsoft:

OMG and Microsoft have been working on this deal since June 2009.

Official Announcement here

OMG contacted me in April 2008 asking if I was interested in working with them, I did wonder why!

15 October 2009

OMG plc

(“OMG” or the “Group”)

IP License Agreement with Microsoft

- Vicon signs license agreement with Microsoft to develop new medical technology -

OMG plc, Oxford Metrics Group (LSE: OMG), (“OMG” or “the Group”) the technology group providing

image understanding products for the entertainment, defence, life science and engineering industries

announced today that it has signed an intellectual property (“IP”) license agreement with Microsoft

Corp.

The agreement will allow the Group’s Vicon arm, which develops motion capture products for the life

science, engineering and entertainment industries, to manufacture and sell devices incorporating

Microsoft’s SenseCam technology worldwide.

SenseCam is a wearable digital camera that takes photographs automatically, without user

intervention. Developed by Microsoft Research Cambridge, the device is most commonly used by

medical researchers as an aid for people with memory loss, but has great potential in a number of

different application areas. Research into these areas will be possible now that it will be more widely

available.

Vicon will unveil a new device, which is based on the existing SenseCam technology on its stand

#1413 at the annual Neuroscience Conference in Chicago on 17-21 October.

Ahead of the device launch OMG plc Chief Executive Officer, Nick Bolton, said:

“I’d like to thank Microsoft and Microsoft Research Cambridge for allowing Vicon to get involved in

such an important product that can support a variety of researchers, including those investigating

memory-loss. It’s an honour to work with Microsoft and to build on the amazing work they have done

with SenseCam over the past five years.

“This IP license agreement means we’re the first company to be able to deliver Microsoft’s ground

breaking technology to a wider audience. All of us at OMG look forward to working closely with

Microsoft and the researchers that have been using SenseCam Project technology and continuing to

move it forward.”

Andrew Herbert, Distinguished Engineer and Managing Director, Microsoft Research

Cambridge said:

“Microsoft’s willingness to license intellectual property arising from its research enables partner

companies to bring innovative products to market. This agreement represents the next stage in the life of SenseCam and will bring the device to a broader community.”

What next after Sensecam? 3D thermal imaging

2009-09-17T04:23:00.000-07:00

What next after SenseCam? Part 2 .

Continuing on from my last post on SenseBulb, people always ask, is the thermal image of a dog or a person in the room? The temperature difference between a dog, cat and human is quite minimal, they do move at different speeds, but what is an easy way to classify living objects from their thermal signatures? Also fridges, outside doors etc have a thermal signature on opening, a step change in temperature followed by about 2 seconds drift to previous room temperature, but are different heights. It seems that as SenseBulb will also have a conventional wall switch positioned half way up wall, it is an ideal place for another array of thermal sensors, to add height (z) data to the x, y of room position.

So I presented my 3D thermal imaging bulb today. I have only built half of it, (the ceiling part) hence image of 4 x4 x4 Rubik's cube as analogy. Thermopiles in ceiling and wall could also provide video animations (avatars) of people sitting, standing, cats, dogs, fridge doors etc. Main research use for assisted living, but could be useful for on 3D control applications, e.g. data input to computer using hand gestures.

The original 2D SenseBulb seems to solve a lot of challenges found with other security devices, it replaces door micro switches, floor mats, energy wasting (a flat screen TV has a high thermal profile, people forget to switch them when leaving room in my house). Passive Infra Red sensors do not respond to static objects, e.g. a seated person. It would also be reassuring just to know where the cat had hidden herself today...

What next after Microsoft's SenseCam? SenseBulb - The Lightbulb that cares

2009-09-04T11:19:00.001-07:00

At Microsoft Reseach Cambridge in 1999 I invented the SenseCam , a sensing camera that captured images of all the minutia of the day. I designed SenseCam to aid people with Alzheimer's and the team at Addenbrookes Hospital , Cambridge and Microsoft did a great job with proving this on patients. One patient could not remember more than 1 day, with SenseCam was increased to 4 months.

I left Microsoft Research Cambridge in 2007 and then set up Girton Labs. I wanted to move forward with my research using novel cameras to help people with memory loss. I was also obsessed with making the lightbulb more useful, we know about low power White Light Emitting Diodes for home lighting, but the light bulb in our ceiling is a bird's eye view to watch over and care for us. I wanted to solve problems like leaving the gas hob on and walking out of the kitchen, and forgetting to come back and so a burnt pan or worse. If we live on our own (like I do) and fall and have an accident, how do we get help? We don't want cameras (too invasive and too much data to process ), might not be able to reach our phone, and don't want gadgets strapped to our body, especially when in the bathroom. The answer - SenseBulb, covered here today by the BBC . Thermopiles measure temperature at a distance, they are used in heat seeking missiles and also non contact ear thermometers. I built some SenseBulb prototypes to measure temperatures of the objects in my kitchen. If we leave the hob on and a person is in the kitchen, that is considered safe, if we walk out of room for 20 mins and leave hob on, an alert needs to be sent. I initially tested bulbs with wireless (433Mhz) messages to the PC and then relayed message via internet. Bu not all people have the internet ( some also switch wireless network off at night!) and so I redesigned it to have a SIM card in lightbulb and send text messages directly to your cell phone or a friend. Messages in test took about 30 secs to be received (we can pay a premium for faster, reliable delivery times). The device, called SenseBulb, used very fast sensors, sampling temperature 200 times second, and this can also detect hand waves. You fall and don't move for a time period, bulb "calls a friend". You fall, but can also wave at bulb, bulb flashes to acknowledge, sends text to a friend, who calls to check up. Bulb also used for normal family care, I am a single parent, I might be out at the pub, but bulb sends me text messages to say "fridge door been opened 4 times in last hour", so I know all is well at home with my family. Detecting a fridge door open doesn't need switches, a thermal image signature from door open (quick temperature drop, then slow rise) is sufficient. Same with the cold draft from the cat flap, I get text message sent when cat in!

The state of the project, Sept 2009, a working PCB of hardware and software, sending GSM messages via Orange mobile phone network. People want to use the device, e.g. elderly people on their own. What next? I'm looking for some investment to cover my time, to build a smaller pcb and bulb, adjust algorithms for more messages, wrapped in a beautiful 3D CAD designed lamp enclosure, make some prototypes and delivery to university and medical researchers. I estimate a bulb could be sold for <$100, business plan is similar to a cell phone, annual contract for 500 messages/ month.

More here

Girton Labs Sensepaper

2009-08-03T00:38:00.000-07:00

I am Lyndsay Williams, Managing Director of Girton Labs, Cambridge, England . We design unique hardware sensing platforms for mobile phones and wearable computers.
Day I of my new blog, a fine sunny day in Cambridge, and I am working on new printing techniques for my sensesurface project.

this is a hybrid paper computer, 2mm thin.

My printing press keeps smoking and blocking up so I am not

bothering to to put all the screws back in as it is taken apart

so often.

Interesting PowerPoint Talk from Microsoft Research Cambridge on

Microsoft and Intellectual Property- pdf file

Talk is using example of "Rita Researcher", her hardware invention and patent on Human Body Networking. Here is my Microsoft Patent on Human Body Networking http://tinyurl.com/22t679. The PowerPoint explains how and why Microsoft work with Intellectual Property. There aren't many Rita's like that at Microsoft... now :-)

Image of SensePaper prototype - DrinkMe