Innovations

Citation

Rudall, B.H. (1999), "Innovations", Kybernetes, Vol. 28 No. 9. https://doi.org/10.1108/k.1999.06728iaa.002

Innovations

Keywords Automation, Cybernetics, Research, Technological developments

Abstacts Reports and surveys are given of selected current research and development in systems and cybernetics. They include: RSI and the brain, Innovations, Biocybernetics, Mathematics and cybernetics, Molecular devices, Devices controlled by thought, Automation and cybernetics, VDU radiation.

Innovations

1. Computer scanning project

A report in the Lancet on the early results from a study called the Early Lung Cancer Action Project showed that computer scans (CT Scans) detected three times more pre-cancerous swellings than chest X-rays could. The findings also included statistics that showed that such scanning had found four times more cancerous nodules and six times more early stage tumours. The CT scans spotted malignant tumours that were substantially smaller than those detected using the current chest X-rays. The project also reported that of the 233 people found to have pre-cancerous swellings 27 went on to have developed malignant tumours.

This indicates the success of using modern computer technology in fighting what is said to be the biggest killer, with fewer than 10 percent of patients still alive five years after diagnosis. The usual conventional chest X-rays have not been able to pick up the signs well enough to be of use to doctors. Using CT scans, the report says, could mean that lung cancer could be detected early enough to save lives.

The study has been initiated by US and Canadian scientists. They have been led by Dr Claudia Henschke (New York Presbyterian Hospital). The technique they have tested is one of computed tomography. This brings together both computer systems and the X-ray devices. The result of the combination is a system that is able to provide clear cross-sectional images of tissue. The initial tests have been on 1,000 voluntary subjects, chosen because they were considered to be at a high risk of lung cancer. The published results in the Lancet of the use of this innovative technique are encouraging but, of course, are based on initial testing, and are at an early stage.

2. Micro-robots

A report from Japan indicates that several Japanese electronics companies are developing an ant-sized robot for applications directed at power plants. This is designed to produce a robot the size of an ant that will be capable of crawling around thin pipes to inspect and repair the system.

The robot is only five millimetres (about 0.2 inches) long and weighs just 0.42g (approximately 0.0147 ounces). The designers claim that the prototype ant-robot is able to lift objects that are twice as heavy as itself.

There is little doubt about the potential for such machines in industry in general but particularly in the nuclear industry. The robots are expected to be in use soon and the companies involved in their development say that they will most certainly speed up inspection and repairs at electrical plants and in some of the nuclear stations.

3. Software to detect lying

The development of systems that are able to detect when a person is lying have been under development for many decades. While the original “lie-detectors” are now in museums, many such primitive devices are still in use. Many government departments and institutions need these methods urgently and after the failures of so many in the past are anxious for systems which show a much greater degree of accuracy. Now in the USA the Salk Institute, in California, have turned to computer technology. They report that their scientists have produced a computer program that they say is capable of:

imitating one of the most delicate and prized of human skills. That is: to tell whether somebody is lying.

The software has been developed to process data that have been captured on thousands of facial movements. Students of the fields psychology and machine interface will already be aware of the research that has been carried out on the analysis of such facial movements. First the technology to record the movements has been perfected and various analysis programmes undertaken. In this instance the Salk researchers are developing techniques that will enable computers to detect the “signs of conflicting emotion that typically accompany lying”. These they claim are facial movements such as “the beginning of a scowl followed by a smile”. This sounds rather simplistic but the scientists claim that their innovative approach has much more success than the conventional lie-detectors.

The software, however, is still in its early stages and we are told that it will not be available for some years.

4. Innovative approach to hard system development problems

A project, funded by the UK’s Engineering and Physical Sciences Research Council (EPSRC) led by Middlesex University, UK, has broken new ground with Programmable User Modelling (PUM) analysis. The university researchers have applied PUM analysis to solve hard system development problems. A report in EPSRC’s Impact (No. 23, 1999) says that working with the UK’s Hertfordshire and York Universities the Middlesex University-led project has now demonstrated the power of PUM “in a practical setting”.

The research file summary of the work says that:

Mathematical models which encapsulate the human dimension needed for the design and safety engineering of interactive systems have been effectively applied to the initial design phase of a live communications system being built by Praxis Critical Systems.

We are told that PUM matches software engineering techniques with formal analyses of task scenarios, operational behaviour, user knowledge and other types of human factors.

Dr Ann Blandford, of the Middlesex University School of Computing Science and the principal investor, says that:

Historically PUM analyses have generally been too complex and time-consuming to be integrated into practical developments that have stringent time and cost constraints.

A User Interface Technical Consultant for Praxis, Sue Milner, believes that this kind of application to a real problem in an industrial setting is a vital step in convincing us that academic research can solve the hard system development problems we deal with everyday.

The report indicates that a key feature of PUM knowledge is the asking of questions such as:

What are the likely consequences of the user not knowing this? It would appear that exploring this line of thought led to the discovery of a number of safety issues, such as those which require users to report “unexpected” or “missing” communications.

The PUM Application (PUMA) project, we are told by Professor Blandford, has the ability to introduce human factor concerns as designers and safety engineers follow their standard procedures as a key goal. PUMA has also investigated the more theoretical aspects of PUM. Dr Butterworth of the PUMA team at Middlesex says that:

We have mathematically encoded the assumptions underlying PUM. In doing so, we have been able to address deep issues concerning the usefulness of rigorously expressed user assumptions in models of interactive systems.

The Web Links to this innovative project are: www.cs.mdx.ac.uk/puma (project) www.praxis-cs.co.uk (Praxis Critical Systems).