Thursday, 13 December 2007

Sleep Research

Helped with initial conceptual design of a system to assist with sleep research in a care home for the elderly.

The system will log the amount of time each individual patient spends inside a particular room emitting a certain wavelength of light. This data will later be correlated to the patient's sleep pattern (measured with a different device) to determine if sleep is affected by the light.

This work is part of PhD research by Lloyd Morgan at the University of Surrey.

Wednesday, 12 December 2007

Enactive Torch recieves funding

The Enactive Torch has recieved 'pump priming funds' from Sussex University's Centre for Research in Cognitive Science (COGS).

These funds will enable the construction of multiple torches and will also permit a more formal and extensive set of experiments with the device.

New Logo

New logos have been designed (using Inkscape) for the artistic adventures of the AdLab. One version can be seen in the Navigation bar on the right hand side of this page.

Watch this space for stickerpack news.

Thursday, 6 December 2007

Do No Harm! (or as little as possible)

Looking at 'human-friendly' robots (i.e. the type that won't crush you against a wall by accident). Found an interesting paper which included the following image:

I can already see the tabloid headline "Robots developed to smash heads!".

The paper is called 'Approaching Asimov’s 1st Law' so someone has a sense of humour!

Tuesday, 4 December 2007

Uncanny Valley

Recently started reading Issac Asimov again. Knowing a bit more about robotics now it's interesting to see that some of the issues he raises in Caves of Steel (of disgust and distrust of an overly lifelike humanoid robot) can be regarded as a precursor to Masahiro Mori's Uncanny Valley. In this idea a designer must be careful not to create a robot that is unnervingly almost real (i.e. by not blinking or being cold to the touch).

I do wonder if I'll ever have to may some of the socio-robotics problems that Asimov is famous for. Of course his sacred first law of robotics (...a robot can never harm a human...) was broken when the US started making Talons.

Thursday, 29 November 2007

New Media on Enactive Torch

Our Enactive '07 Oral presentation has been added to the Enactive Torch Website.

Go direct to the Audio stream by clicking here

To view the presented paper: 'Towards a Phenomenological Pragmatics of Enactive Perception' click here

Wednesday, 28 November 2007

Enactive '07

Recently returned from the Enactive Interfaces / Enaction in Arts conference in Grenoble (France) where me and Tom presented our paper on the Enactive Torch. Met some very interesting people and our work was well recieved.

Also at the conference was Peter Bennet who also studied Cybernetics at Reading with me and Tom.

Tuesday, 13 November 2007

Enactive Torch Experiments

Experiments are planned for this Friday with the improved Enactive torch.

A logo has also been designed. The left side of the image shows the Braille representation of 'ET'.

Sunday, 11 November 2007

Enactive Torch Website

The Enactive Torch now has a dedicated website:

As well as being generally informative the device aims to provide a point of contact for researchers who may be interested in using the enactive torch in their own studies of perception and sensory substitution.

Thursday, 8 November 2007

Vicon Motion Capture

Have been looking into the Vicon Motion Capture System I'll be using. Looks like its going to be fun, interesting and then very difficult (if their software doesn't do what we want).

Enactive Torch phase 2

Work on the enactive torch modifications is going well the problem of oscillations between readings now fixed. This has been a consistant problem until now and it still requires some work. In the enactive torch these oscillations have severly impaired the creation of the extra sense as the environment is suddenly uncertain. This is also true in the haptic torch were I was previously informed (by a blind person) that this was a serious bug that would stop them using the system.

I'm hoping to get some trials recorded this weekend but as always the problem lies in somehow getting people to an unfamiliar environment.

Wednesday, 7 November 2007

Warwick Lecture

Went to see a public lecture by my old boss Kevin Warwick yesteday. As opposed to talking about Robots taking over the world he's now talking about the implant projects and placing mixed raw neural tissue (from unborn rats) onto microelectrode arrays to create biological neural networks. It all sounds very sci-fi and the audience loved it but to be honest neurophysiologists have been doing MEA work like this for decades so I've yet to be sold.

At the end some undergraduates were asking Kevin an implant question which he handed over to me. So now I'm helping these guys with their project (and I thought I'd left the field)

Thursday, 1 November 2007

PhD Started

The Will's Memorial
Have started my PhD at The University of Bristol.

It appears that I'll be dividing my time between the Control Lab at the Department of Mechanical Engineering and the Bristol Robotics Lab.

Wednesday, 17 October 2007

Goodbye Reactor

The Boom

The Reactor - note robot on right

Today is my last day working on Remote Handling at the Joint European Torus (JET). I've certainly learned alot about large scale applied robotics in this very unique environment. The size, precision, man-in-the-loop feedback, Human-machine-interfaces and large forces involved in the systems here are much greater than anything I'd come across while working at a University (something to do with funding perhaps).

In the words of my current boss 'they struggeled to keep me stimulated' which is why its onwards to the PhD...

Interestingly the anthropomorphic robot here (you can see it with its arms folded in the above image) has a limited range of abilities (though much greater than any robot I'd seen elsewhere) so tools are specially designed and adapted to be used by this system. In my PhD I'll be doing the opposite and attempting to get a humanoid robot to move like a human and therefore be able to use human tools.

Mail Strike Cancelled

New sensor finally here! Seems to work but analogue voltage response appears to be non-linear, a little bit annoying but shouldn't take that long to figure out.

I have 1 evening to integrate it into the enactive torch before moving out of my house and into desk-less limbo. The race is on!

Tuesday, 16 October 2007

Another good definition

"Enactive knowledge is neither symbolic nor iconic. It is direct, in the sense that it is natural and intuitive, based on experience and the perceptual consequences of motor acts" [from the Enactive07 website.]

Friday, 12 October 2007

Enactive Knowledge

Found a great definition of Enactive Knowledge.

I'm still learning about Enaction in preparation for the upcoming Enactive 07 conference. It appears that such devices as the Haptic Torch are Enactive Systems due to the physical motion used by the user to identify scenery. I've always thought this to be better than a system that simply presents all obstacle information to the user without any sort of feedback. All you end up with is redundant and unwanted information that confuses the user and makes it harder to identify important features.

New Enactive Rig

Dreamed up some improvements to the Enactive Rig and modelled it in Catia. Looks like it should work. Time and money are of course the deciding factors. Would hate to spend money on getting this thing manufactured then leave it sitting on a desk because I'm too busy with PhD.

Need to get priorities right that's for sure!

Wednesday, 10 October 2007

DS Homebrew

Discovered this great gadget awhile ago (but quiet at work today so I discovered it a bit more). Its a RoboDS which seems to be the pre-built application of the DSerial2 cart.

Was thinking of getting into arduino as PICs are a pain to source and the MPLAB environment isn't as friendly as processing. This might sway me though as the DS would make a great development platform (its already got sound, a screen for debugging and a touch screen for an interface!) Not sure how much memory you have to play with but an SD adaptor in the other (front) game slot could solve this for $40 (and I already have a DS - bonus!).

Of course the temptation to play Zelda when things are getting bad would require some hefty will power. It would also be rather gutting to see both screens permenantly blank out after accidently pouring in 10V.

Tuesday, 9 October 2007

Enactive Rig

Spent a bit of time designing a (very simple) desk based enactive torch rig.

Really this model just illustrates a concept (and is better than trying to describe such a system verbally).

Thursday, 4 October 2007

Simple Life Form

Got the animation studio set up again last night, its been awhile but made this (and a few other things). Am still just playing with this medium before commiting to something bigger (as you can imagine it takes ages to make anything). Also still playing with different software, this one is AnimatorDV, pretty good for free.

Royal Mail halts science

About to start working with a new sensor for the enactive torch. Looks great, sounds even better. Where is it? Ask her majesty's postmen! This is the second time a strike has got in the way of me and this sensor.
This new ultrasound should get rid of the some of the glitches that plauged the haptic torch. The main one being a dead-zone which ignored anything closer than about 2 inches.
With any luck this post strike (the third one this summer!) will be over in less than a week...


Welcome to The_Ad_Lab! I tried setting this up before but didn't really keep up with it, however I'm about to leave my day job, PhD start date is getting closer and some good things are happening in my other research. Lets start blogging!