Future trends in cybernetics
Keywords:Automation, Cybernetics, Research, Technological developments
Abstract:Gives reports and surveys of selected research and development in systems and cybernetics. They include: Future Trends in Cybernetics; Imaging Process; Telemedicine; Information Systems; Cybernetics and Robotics; New Research Activities developments.
Future trends in cybernetics
What is Cybernetics?
Future trends in cybernetics and systems are, of course, hard to predict, but the wide range of papers published in this journal and in the community at large can only reinforce our faith in the inter- and transdisciplinary approaches the fields embody and which are characteristic of these endeavours.
One of the main problems highlighted by Professor Stafford Beer in his address published in this issue, is What is Cybernetics? He says that most people have no more than a hazy idea of what it is all about. He says "I have often been assured that it is about freezing people – but they were thinking of cryogenics" Readers and their colleagues in academe and business should be encouraged to read Stafford Beer's address which shows not only how novel the approaches made in cybernetics and systems have become but also what their relevance is to what is called real-life.
Identity crisis
Feedback from cyberneticians and systemists who have participated in recent conferences and meetings worldwide also suggest that in our current "high- tech age", cybernetics, is suffering from an identity crisis. The difficulty appears to stem from what these fields encompass. There are also concerns that not only,as Professor Beer suggests, is there ignorance about what cybernetics is, its very name has been "hijacked" by the media and indeed any company or organisation that wanted a modern-sounding title and hightech-image to match. Product names and services in particular now frequently use the prefix cyber in the marketplace. Several uses are now firmly accepted by the public in connection with the Internet and words such as cyberspace, defined as meaning "a notional environment in which electronic communication occurs or virtual reality". How many more names have been constructed with cyber as their prefix?The consequence of their use, we are told tends to cloud the meaning of cybernetics. The history of cybernetics in the last centuries chronicles the attempts of many scholars to describe and define its meaning. The definition"the science of communications and automatic control systems in both machines and living things" used so often in the literature seems inadequate in the 21st Century. It would appear that the many central ideas that are typical of the cybernetical attitude and which carry with it something more than an appreciation of control by informational feedback are now universally promoted,as can be judged by this current volume of Kybernetes.
Readers are reminded that the section of this journal Communications and forum welcomes comments and discussions on this and similar topics. The cybernetics community will benefit from the viewpoints of readers on the important matters concerned with cybernetics and systems that arise in the contributions published here and your feedback is invited and positively encouraged.
Imaging processing
Processing images and their interpretation
A report from the University of Manchester, UK describes how their Imaging Science and Biomedical Engineering Department have developed a unified approach to face image processing.
A large part of the human visual cortex ,we are told, is devoted to processing face images. Indeed, this reflects the fact that faces play a central role in the way we understand and interact with the world. Much of our time is taken with identifing individuals, inferring their intentions, and augumenting verbal communication.
This section has reported on many occasions the developments in the use of automated face recognition in such applications as access control surveillance,criminal investigation, and coding. It has been obvious that existing technology is limited and that there is a need for systems to be developed that are more flexible and robust.
Creating a natural human-computer interface
The creation of a natural human computer interface has been the goal of computer scientists since multi-access systems were introduced. In many ways, as most cyberneticians will accept, using a computer is still a very primitive operation. Keying-in data, for example can hardly be regarded as a natural or efficient way of interfacing with a computing machine. Any process that makes interaction with a computer system more natural is to be welcomed. Obviously efficient facial recognition systems have an important. part to play in building natural interfaces.
In the long term it is believed that developing capabilities approaching those of humans will be the key to creating natural human-computer interfaces.This research report informs us that:
Automated interpretation is difficult because the scenes of interest are complex and highly variable. Also, most applications of interest involve the interpretation of video, rather than static images.The functionality required to tackle a broad range of applications can be expressed in terms of a number of generic capabilities: feature location and tracking, person identification, expression recognition, pose recovery, coding, and so on. It has been common to treat these as separate problems, but this is naive since, in practice, they are interlinked. For example, a fully functional system for person identification needs to "understand" expression change in order to ignore irrelevant within-person variation.
The key, we are told, to the approach taken at Manchester is to use a generative model of face appearance that takes into account all the main sources of variablity. Professor Chris Taylor and Dr Tim Cootes who have developed this new unified approach say that:
By varying a set of model parameters it is possible to generate a synthetic image of any person in any pose with any expression and so on. The model deals with variability in both shape and intensity and, because it would be very difficult to "hand craft", it is learned by applying statistical analysis to a large set of training images. This results in a model with around 100 parameters, capable of generating photo-realistic face images demonstrating a full range of variability.
Further statisical analvsis
Further statisical analysis, we are told, can be used to partition the parameters into subsets that each control exclusively one aspect of variability-identity, expression, etc. As a spin-off of the work these models can be used directly in computer games and avatars.
Professor Taylor says that:
Given a parameterised model of a face appearance, interpretation of an image containing a face can be posed as finding the set of model parameters that generate a synthetic face as similar as possible to that image. This is a difficult optimisation problem but an efficient solution can be found by exploiting the insight that a very similar problem is solved every time the face model is matched to an image containing a face. This allows the characteristics of the search space to be learned off-line, leading to a robust, or near real-time matching algorithm.
The researchers say that once the model has been matched, its parameters summarise everything there is to know about the image. Because they are partitioned into subsets that deal with specific sources of viariability the parameters can be used directly to identify the individual, recognise their expression, determine their gaze point, and so on.
Active appearance model
In summary Professor Taylor says that:
This approach has been extended to deal with video sequences by modelling the dynamics of the face model parameters, noting that whilst expression, pose and other characteristics can vary during a sequence, identity must remain constant– in real life if not in the movies. One consequence of this approach is that identity information can be integrated over a sequence, allowing a good estimate of appearance to be obtained even from poor quality "Crimewatch"videos.
It is claimed that this "Active Appearance Model" approach has been widely recognised as a significant breakthrough and is currently being applied to a range of challenging applications using UK Department of Trade and Industry(DTI) Link, European Union (EU) and commercial funding. These applications can be as diverse, we are told, as printed circuit board, inspection and medical image analysis.
Further information can be obtained from the Manchester University website: http://www.isbe.man.ac.uk
Telemedicine
Revolutionary procedure
The revolutionary procedure which is claimed to be the world's first example of intercontinental surgery and which is said to shatter distance records for the technique is called Operation Lindbergh. The team of Franco-Americans used"virtual scalpels" linking them to a computer in New York to control the movements of Zeus, a robot in the operating theatre of the Strasbourg hospital. This machine replicated their hand strokes precisely. We are told by the team that the electronic commands and video footage made the round trip in 155 milliseconds and the actual operation on MmeSchall, the French patient, took less than hour to complete. Further details are given in the following sections.
Details of the surgery were published in Nature (September 2001), but we are told that on September 7, 2001, it was performed by Professor Marescaux and his colleague, Michel Gagner using a computer console at Mount Sinai medical Centre in New York. Video footage of each scapel movement was relayed to the doctors with, it is claimed, a time delay too short to be noticed by the human brain.
The Zeus equipment took 16 minutes to set up, and the procedure was completed in 54 minutes. Every aspect, bar the initial incision and post-operative stitching was performed by the robot. Since the procedure was in a sense still experimental a full team of surgeons were standing by in case of any problems.
The future of remote surgery
The success of Operation Lindbergh heralds a new era in which surgeons will be able to operate on patients anywhere in the world without leaving their own theatres. What it means is that access to specialised skills and training will be completely transformed. Life-saving surgery by robots within metres of the battlefield will be possible for military personnel wounded in combat or in accidents, all at a minimal risk to the doctors involved in the procedure. Professor Marescaux of the Louis Pasteur University of Strasbourg, the leader of the surgical team believes that:
It was a milestone in medical history. We are looking at a technique that will make surgery completely global.
A transatlantic operation
Most developers are now afraid of using "the first ever" caption for their endeavours whilst manufacturers have no qualms about making such claims. The recent report received about this transatlantic operation to remove a gall bladder proudly announced it to be the "first transatlantic operation". It said that surgeons in New York carried out the operation using remote controlled robots to remove the gall bladder of a woman in France.
The procedure was carried out by three French surgeons at a control console equipped with monitor screens. We are told that the careful and delicate movements of the surgeons hands in New York were electronically transmitted a distance of 4,300 miles to an operating theatre in Strasbourg, where a set of robot arms obeyed their commands.
The result was that the gall bladder was successfully removed in just under an hour and we are informed that the 68-year old patient was discharged from Strasbourg Civil Hospital 48 hours later.
It is important to note the difference between this reported remote controlled operation and some others that have been held. Many "telemedicine"operations have been confined to providing communication links and a limited degree of surgical expertise. In this transatlantic operation it is claimed that for the first time surgeons have completed a long distance remote-controlled operation themselves.
One of the problems that such operations encounter is the "time lag" which is, of course, caused by transmitting the information along the telecommunications lines or other linking system. In this case we are told a"high-speed optical network" was used in an attempt to increase the speeds of transmission of the data required by the remote robot-surgeons that are actually in contact with the patient being operated upon. Although advances in both communication, robotic devices and the visual interface between them and their surgeon-operators continue there will at present be some reluctance to use such systems for everyday routine surgery.
Although, as we will see in the detailed account of the Franco-American Transatlantic operation, the time delay in transmitting data for this particular surgical procedure was not a drawback. He also believes that similar methods could be applied to many other complicated treatments. For example, he said:
there are few things for which this is impractical. I am convinced that in less than a year we are going to see the first coronary bypass operation by telesurgery, and we will also see it being applied to other areas such as vascular and gynaecological surgery.
The views of David Rosin, a consultant surgeon at the UKs St Mary's Hospital,Paddington, London, were given to the Times newspaper (20.09.01). He took the view that:
For routine surgery, it will probably remain largely unnecessary, but for extremely specialised surgery which can be undertaken by very few surgeons throughout the world, then this is an absolute boon.
Even so, as with most advances in medicine that are related to the progess of high technology, it is difficult to foresee many of the "spin-off" applications and uses. It could, for example encourage the establishment of centres of surgical expertise which will make its services available both nationally and globally. In consequence we might see the hospitals as we know them change beyond recognition.
Information systems
"Natural" evolution for business
The Research File of the UK's Information Technology and Computer Science Update – IMPACT (Issue 30, Sept. 2001) reports that "Natural" evolution for business can be taught. It suggests that Companies should be able to organise themselves in a way that allows them to continuously adapt to a changing commercial environment. If an organsiation, it says, is sufficiently fluid, it can reduce the periodic need for expensive and disruptive restructuring which often ends in failure.
Integrated complex social systems project
One of the United Kingdom's leading Information Systems Department, at theLondon School of Economics (LSE), has been awarded a major three-year grant to work with Shell Internet Works, Rolls Royce Marine and BT Brightstar (BTs new business incubator) to investigate ways of introducing this "change" concept into a range of different business settings.
The project, called the Integrated Complex Social Systems (ICoSS)project has received funding from the UKs Engineering and Physical Sciences Research Council (EPSRC). The ideas for the project, we are told, come from Eve Mitleton-Kelly, the LSE's Director of Complexity and Organisational Learning Research Programme. In addition a multi-disciplinary group of international advisors from academe and business will also contribute different perspectives,knowledge and expertise to the project.
The key to the work, it is reported, lies in "complex adaptive Systems" which are described as:
A complex system is one in which the system's many constituent components are intricately inter-related and a change in one element causes changes throughout the system in ways that are difficult to predict. In many biological systems,such complexity enables the system to adapt to changes in its environment– hence complex adaptive systems.
Eve Miitleton-Kelly says that:
"The fact that these systems can change and adapt is crucial," ... "If we can understand how organisations of humans function as complex adaptive systems we can then change the way we think about them and how we manage organisations."
One of the most important characteristics of a complex adaptive system is that it can self-organise, rearranging itself in response to changing circumstamces. Cyberneticians will, of course, recognise such systems.
An example of the resulting creativity
An example of the sorts of creativity that can arise in this way has already been observed in earlier researches. A global banking company had a tight deadline to implement a new strategy to deal with the introduction of the new European Currency, the Euro. This required urgent updating and reconfiguring of the company's entire computer system. Dr Mitleton-Kelly describes the circumstances and the results:
"Quite spontaneously one of the local project managers did something which made all the difference,""He invited all the business project managers and the systems developers to meet each other one afternoon on a regular basis. These are people who do not normally feel comfortable talking to one another and were initially not very happy to do this. But gradually they started making connections with each other. The banking people began to understand the problems and issues faced by the computer people, and vice versa. This simple opening up of channels helped get the task done efficiently and quickly. The social interactions provided solutions to the technical problems."
The researchers are aware that all organisations consisting of humans are necessarily complex and that the key is to exploit the advantages of complexity,the ability to adapt and self organise.
Dr Mitleton-Kelly believes that:
"The main thing is to identify the barriers that allow the beneficial aspects of complexity to emerge, and then remove those barriers. Traditionally many managerial approaches have served to stifle these benefits of complexity. What happened in the banking example was that people were allowed to self-organise,to create new relationships."
Adaptive complexity.
It is by working closely with companies that the research team will be able to examine how adaptive complexity might be allowed to flourish in a situation where two companies are merging, where there is restructuring, and in an"incubator" – an "ideas factory" – where employees are encouraged to dream up innovative ideas for new business.
Mitleton-Kelly says that:
"One of the big advantages of allowing "complexity thinking" to flourish is that you should no longer need constant interventions to impose change," "People themselves should be able to change with the external environment because of the inherent fluidity and adaptability of the organisation."
The research team report that they hope to identify some generic conditions or enabling frameworks that will help an organisation to create new ways of working and relating, as well as frameworks specific to the companies collaborating in the project. The main aim is to try to change the way of thinking about an organisation. They believe that it is a conceptual framework,not a case of laying down strict rules or coming up with a particular recipe.
The programme is active and was launched last year when 30 participants including representatives from all the industrial partners as well as advisors met at a meeting opened by the Director of the London School of Economics Professor Tony Giddens.
Contacts
More details and further contact can be made with: Eve Mitleton-Kelly on (UK)Tel: 020 7635 5553. E-mail: mitleton-kelly@Ise.ac.ukor Slavica Savic –s.savic@lse.u
Cybernetics and Robotics
- 1.
Designing a robot dog
Many cyberneticians and systemists will ask why anyone should spend their time and resources designing a robotic dog. Several companies, however, see the project simply as a design challenge and an opportunity to demonstrate the use of advanced software in what has to be a collaborative endeavour. The design team for such a robot dog project was dispersed around different locations and needed advanced software to support a collaborative development environment that allowed complex design information to be shared over the Internet ... The result was what is claimed to be the world's most advanced autonomous legged robot,developed and marked as the RS-01. RoboDog.
Produced with the aid of collaborative design software it is being marketed after less than a year's development, and was available from April 2001 after its launch by its British developers RoboScience. The company say that it is:
Intended to be both an advanced robotics technology demonstrator and a limited-edition commercial product, the RS-01 RoboDog incorporates breakthrough technology in a number of areas. This technology will form the platform for the next-generation lightweight robotics required for automating a wide range of industrial and domestic tasks and for eliminating the need for human involvement in high-risk industrial and military environments.
The size of an adult Labrador and powerful enough to lift itself up carrying a five-year old child on its back, the RS-01 RoboDog was developed in just seven months – from initial design concept to finished, working product –thanks to an Internet-based collaborative engineering environment centred on the UGS* product development and manufacturing software suite, Unigraphics.
Two of the most significant design and technology breakthroughs made during the development of the RoboDog were in the areas of its joints and in the body construction, where the use of the Unigraphics software suite was critical to the project's success. The RoboScience.developing team report that:
"The joints that we have developed can be thought of as the robot's muscles,combining elements that are usually separated, such as hinges, motors and a gearbox in a way that provides a far superior power-to-weight ratio than conventional designs. This is one of the secrets of the RS-01 RoboDog's advanced motive abilities and run time."
During the design of the joints, design data from an external design spreadsheet was imported into the software, where it was used in creating the basic geometry for the parametric 3D models. The final detailed design and assembly modelling of the joints was then completed in the usual way in the 3D CAD environment. Using Unigraphics, the design work and the production of manufacturing information was able to be completed in a matter of days, with the result that working prototypes of the joints were ready for testing less than a month after the design work had begun.
Meanwhile, the body presented another, altogether different design challenge. The RS-01 RoboDog body is an exoskeleton, or monocoque construction. So in fact,it is more like a crab or a scorpion than a dog. The exoskeleton construction presented two main difficulties:
The first was that, because we were designing it from the outside in, there was a chance that we could end up without enough space to fit everything inside properly. The second was that with an exoskeleton, the styling parts are the same as the structural parts and are always being adjusted during the design process. In fact, with RoboDog, a greater percentage of the structural parts are visible than on an aircraft or car.
A further important factor was that, because of the time constraints under which the development team was working, design of the body construction had to begin before the joints and other internal components had been fully tested and finalised.
Throughout the design of the RS-01 RoboDog, particularly when it came to its body, UGS' UG/Visualise software – part of the company's Shape Studio industrial design software suite which fully integrates with Unigraphics –was used to create photorealistic visualisations for styling decisions and for use in future marketing activities.
RoboScience believes the development would not have been possible without the use of advanced collaborative design software, particularly in a tight development environment. This resulted in their ability to integrate the different processes in a single 3D mode1 and communicate them over the Internet.
Apart from the operating system used in its robots (Microsoft™ Windows™)and some other secondary software components, RoboScience owns all of the intellectual property rights (IPR) and patents that are being pursued as a result of the development of the RS-01 RoboDog.
Further details: Contact: RoboScience: nick.wirth@roboscience.com and UGS(Software/Services): kate.mills@ugs.com
- 2.
Automation survey
A survey by the United Nations Economic Commission for Europe (UN/ECE)and the International Federation of Robotics (IFR) brings encouragement to automation in Europe. Orders, we are told, for industrial robots indicate a boom. Whether this can be maintained in view of the threatened downturn in business remains to be seen.
"Never before have so many orders for industrial robots been placed by the European industry, pointing towards an acceleration in the drive to automate",says Jan Karlsson of the United Nations Economic Commission for Europe (UN/ECE),which together with the International Federation of Robotics (IFR), regularly survey the robot market. In 2000, orders for industrial robots in Europe were 25 per cent higher than in 1999. Judging by the fourth quarter of 2000, which showed an increase of 24 per cent, there is no slow-down in sight in the European investment activities, at least not as concerns robotics systems.
The figure for Europe is in particular impressive in the light of a 31 per cent surge in 1998 followed by a 12 per cent growth in 1999.
World-wide orders for industrial robots were up 15 per cent in 2000 compared with 1999, slightly down from a growth rate of 20 per cent in 1999 over 1998. The world-wide increase was concentrated to Europe, up 25 per cent, and Asia, up 32 per cent.
After an impressive increase in orders of 60 per cent in 1999 over 1998, it was expected that orders would drop in North America. A fall as high as 27 per cent, however, is a clear indication of the slow-down in the American economy.
Normally, it is the automotive industry that leads the drive to robotize. This was the case in 1999. In 2000, however, there was a significant turn around. World-wide non-automotive industries increased their orders with close to 40 per cent while final automotive assembly only increased by 3 per cent and automotive components with 9 per cent.
In Asia and Europe, non-automotive industries increased their orders with as much as 56 per cent and 41 per cent, respectively. In North America a modest growth of 5 per cent was recorded in orders from the non-automotive industries,in contrast to the automotive industry which recorded sharply falling robot orders.
The surge in robot investments has several explanations. A central reason is that prices of robots are falling rapidly relative to labour costs. Robot prices in year 2000 were on average 44 per cent lower than in 1990. At the same time today's robots have much higher performance than those produced in 1990 with respect to versatility, speed, accuracy, and above all computer power.
A quality adjusted price index, that is an index in which consideration is taken to the continuous performance improvements, would therefore show an even higher price reduction. It is estimated that a robot in year 2000 costs less than one third of a robot in 1990 with the same performance.
While prices of robots have plummeted and quality has increased, labour costs have risen steadily. In the United States, for instance, labour compensation in the business sector rose by 43 per cent in the period 1990-2000. In the same time robot prices in the US dropped by over 60 per cent without taking any consideration to improved robot qualities and performance. By taking the later into account prices would have dropped more than 80 per cent showing that robots have had a price/performance development similar to that of personal computers. For every year, robots are becoming more and more cost-effective vis-à-vismanual methods of production.
"Rapidly falling relative prices of robots paired with shrinking labour supply, in particular to the manufacturing sector, in the years ahead will spur continued high growth in robot investment", says Mike Wilson, chairman of IFR."We have only seen the first phase of the robotization drive, which has mainly focused on the automotive industry. The food industry and all other manufacturing industries as well as many non-manufacturing sectors, are as the year 2000 data clearly show, significantly stepping up their investment in robot systems" says Jan Karlsson, UN/ECE.
For more information about the ECE/IFR Survey please contact:
Mr. Jan Karlsson, Statistical Division, United Nations Economic Commission for Europe (UN/ECE), Palais des Nations, CH–1211 Geneva 10, Switzerland. Phone: +(4122) 917 32 85; Fax: +(4122) 917 00 40; e-mail:jan.karlsson@unece.org
International Federation of Robotics (IFR), Box 5510, S–11485 Stockholm, Sweden. Phone: +(46 8) 7820843; Fax: +(46 8) 6603378; e-mail: ifr@vi.se
Mr. Mike Wilson, Chairman of IFR, Meta Vision Systems Ltd., Oakfield House,Oakfield Industrial Estate, Eynsham, Oxfordshire OX8 1TH, United Kingdom. Phone:+(44 1865) 887 900; Fax: +(44 1865) 887 901; e-mail: mike.wilson@metamvs.co.uk
New research activities
- 1.
Finding out how our visual systems and brains work
A report from Newcastle University, UK gives details of the role of binocular eye movements in the perception of motion in depth by human observers. The report says that:
When objects move in the natural world, we usually try to follow them with our eyes. The Newcastle team is currently testing how well and how often people make binocular- following eye movements, and to what extent these eye movements help, or hinder, the visual system when detecting how objects move in depth.Recent research has shown that motion in depth is detected poorly when compared with the detection of lateral motion (when the eyes are fixed). But it is not known whether the results of such work can be generalised to situations when the eyes move, or when the observer moves to intercept an object.The team's next series of experiments will test whether binocular eye movements can improve the detection of motion in depth. Then, it will test whether such movements allow observers to make more accurate estimates of motion in depth, and whether they help people interact with moving objects.
The work of the Newcastle University research team was written by Dr Julie Harris who can be contacted on email: jharris@ncl.ac.uk
- 2.
Novel laser techniques to manipulate chromosomes
A multidisciplinary team of researchers at the University of St. Andrews. Scotland, UK is developing novel laser techniques to manipulate chromosomes with unprecidented precision. The aim the team reports, is to provide biologists with a powerful tool to study the fundamental mechanisms of how chromosomes become damaged.
The research is being led by Dr Kishan Dholakia of the lasers and optics group in the university's School of Physics, in collaboration with Professor Wilson Sibbet and Dr. Peter Bryant also of the university. The programme of research is funded by the UK's Life Science Interface Programme. The researchers say:
The genetic material of all organisms is constantly subjected to attack by environmental factors such as toxins or radiation. If damage does occur, special enzymes are despatched to effect a repair. However, occasionally a break in the strand of DNA can then manifest itself as a break in the chromosome as a whole. It is thought that at least one type of "chromatid" break results from a rearrangement (involving cutting and rejoining sections of the chromosome) of the chromosomal material around the break site, which the cell carries out in response to the original DNA damage. If this rearrangement process is incomplete when the cell divides, a break will be observed when the chromosomes are viewed under the microscope. To study this phenomenon biologists need to isolate damaged chromosomes and then excise the small portion containing the fracture for further analysis. This is a difficult manipulation and it is here that the optical physicists believe they can help.
The researchers say that most biologists currently use thin glass needles to cut small sections from the chromosomes under the microscope. They believe that this is not very sensitive and requires some quite complex mechanical instrumentation. The St. Andrews team is developing a way of using laser light in the form of optical tweezers to move and cut the chromosomes. They say light has a small but definable momentum, and the high power per unit area of a laser beam is capable of producing a force of the order of pico Newtons.
Dr Dholakia says that "These optical tweezers make it possible to pick up and move micron-sized particles".
The researchers reports that:
The wavelength of the light must be selected with care so that the object is not damaged by the intense beam. Once the chromosome of interest has been isolated by the optical tweezer it is necessary to cut around the break to enable that region to be analysed. "To do this we hope to use a second, pulsed,laser with a different wavelength to act as an optical knife." "By manoeuvring this laser around the region of the break we should be able to isolate that region and then tweeze it out."
The St. Andrews researchers also say that they are developing ways of making chromosomes more mobile, and that they have succeeded in manipulating whole chromosomes with the optical tweezing system. Further experimentation with different configurations of laser light to find the most efficient tweezing geometry is being pursued. The researchers believe that once their system has been shown to work for chromosomes they would expect it to be applicable across the life sciences.
- 3.
Minority languages and it developments
A research report from Dr. Tony McEnery of the Department of linguistics and Modern English Literature at the University of Lancaster U.K. outlines the work being carried out in what is described as "minority language engineering". The researchers state the problem clearly by asking you:
to imagine that you are asked to type in English at a keyboard which contains a set of characters that are utterly alien to you. Then imagine that you are asked to surf the web, but before reading any page you have to download and install special files almost every time you access a new site. Finally, imagine that you take a word processor file from your friend and that when you read it instead of seeing English you see alphabet soup.
These are the types of problems encountered frequently by speakers of South Asian languages across the world. The Lancaster University (IT) researchers say that the provision of Information Technology package and basic standards for the encoding of documents in South Asian languages is very much lagging behind what is possible with European languages.
In the UK research funding has been made available by the research council for engineering and physical sciences for a project for Minority Language Engineering (MILLE), in recognition of the importance of these languages in the UK and globally.
MILLE is a project set at Lancaster University which attempts to outline needs for language processing and IT support for South Asian Languages. The project researchers report that:
MILLE showed that while language processing efforts may be currently directed towards European languages, the reality of the matter is that most translation undertaken in the UK is not into European languages – it is into languages associated with immigrant communities in the UK, notably South Asian languages.MILLE established the framework for a new project based at Lancaster and Sheffield Universities – Enabling Minority Language Engineering (EMILLE)– which is seeking to "boot strap" work on language processing for Bangla,Gujarati, Hindi, Panjabi, Singhalese, Tarnil and Urdu. While the project is dealing with some very basic issues focused around the encoding of electronic text and the role Unicode has to play in language processing software development, work is also being undertaken on the development of software that may be of use to such user groups as translation agencies.
The project is making progress in a number of directions. It is reported that:
One area in which work is ongoing is in the area of paralleI alignment– automatically deducting which sentences are translations of which sentences given a set of bilingual texts. Such techniques have proved crucial in making machine translation a viable proposition for a whole range of European languages, and those techniques are now being extended to South Asian llanguages.
It is claimed that by the end of the project an infrastructure for enabling the development of a language processing systems and usable products for South Asian languages will have been developed.
B. H. RudallNorbert Wiener Institute and the University of Wales (UK)
