Common Sense, Scientific Method, and Accident Prevention,
by Professor Cyril Domb
(Emeritus Professor of Physics, Bar Ilan University, and Director of the Nebenzahl Institute for Human Safety and Accident Prevention)

Lecture delivered to the Israel Physical Society at its annual meeting in Bar Ilan University, April 10, 1995

1. Introduction
2. Rare and Horrendous Accidents
3. Facts and Figures
4. Road Accidents
5. Industrial Safety
6. Home Safety
7. Natural Disasters
8. Conclusions

1. Introduction
It is a privilege and pleasure to address the Israeli Physical Society at its Annual Meeting, and, although the subject of my address is not in the area of physics, I hope that what I say will be of interest to physicists.

Let me begin with a personal reminiscence. When I joined King's College London as Professor of Theoretical Physics in 1954 the Head of the Department, the Wheatstone Professor, was Sir John Randall. His immediate predecessor had been Sir Edward Appleton who received a Nobel prize for his work exploring the ionosphere. He pioneered the cathode ray oscilloscope techniques which were later of such importance in the development of radar. Like many top research scientists, Appleton faced the conflict in priorities between advancing his research and administering his department. Appleton was a very practical man, and soon formulated a policy to guide his activities. "There's nothing to administration provided that it is treated seriously." It is hopeless to try to administer properly in a few hours here and there which have been freed from research. The problems to be tackled must be clearly outlined, and given their own appropriate allocation of time.

I should like to adapt Appleton's slogan to the problem which I am about to discuss, and say, "There's nothing to accident prevention provided that it is treated seriously." Bur here "treated seriously" means using scientific methods in analyzing the phenomena which give rise to accidents, and applying common sense where necessary. Common sense should have overall priority, and may occasionally lead to conclusions different from those derived from the use of standard scientific criteria. There will be little progress as long as there is a widespread public illusion that accidents are inevitable and are not governed by normal laws of nature. An article in "Physics Today" in March 1987 tells of accident litigation in which a physicist A. C. Damask told the jury of some experimental work which he had done to throw light on the cause of the accident. In his response the opposing attorney said, "the laws of physics are obeyed in the laboratory but not in rural New Jersey," and this argument convinced the jury! A campaign of public education is essential at all levels to provide a reasoned approach to accidents and their cause, to furnish statistical information and explain its meaning, and to outline a coherent policy aimed at reducing accidents. I believe that a target of a 50% reduction of accidents during the next few years is feasible provided that an appropriate programme of this kind of action is carried out.

2. Rare and Horrendous Accidents
How did I, a physicist, get involved in accident prevention? In December 1971 I traveled with my family from Britain to Israel to take up an appointment as Visiting Professor at the Hebrew University. On the morning of our arrival we heard news of a disaster at a railway intersection in which our best friends in Israel, Rabbi Shlomo Kook, his wife Yehudit, their two youngest children, and a cousin had been killed. The crash took place on the Hulda road during their return journey to Rehovot from Jerusalem. It was a rainy night with poor visibility and they must have failed to identify the exact location of the rails. The car in which they were traveling was hit by an oncoming train with the consequences listed above. Had there been a barrier at the intersection it is almost certain that the tragedy would not have occurred.

How does one react to such a disaster? Of course one can assess the consequences, the loss to Israel of people of extraordinary capability, the shattering of a wonderful family. But far more important is to do everything humanly possible to ensure that such a tragedy is not repeated , and in fairness to the authorities, steps were taken to cut down trees at this intersection, and improve visibility in the direction of oncoming trains. But the crucial step of inserting a barrier at this and all other railway intersection in the country was not taken. Presumably a cost-effective argument was used. Such a collision is extremely rare, and if the expectation value of the damage is calculated, it does not justify the large expenditure involved in providing barriers.

This is the conventional scientific analysis. But I would like to draw attention to an alternative common-sense view that events with horrendous consequences should be avoided, as long as the cost is bearable. This view has the support of the distinguished astrophysicist Sir Hermann Bondi, who persuaded the British government to build a barrier in the river Thames costing 60 million pounds to prevent possible damage to London from an extraordinary flood. The horrendous consequences of such a flood should be avoided, even though it might be expected to occur only once every hundred or two hundred years. Bondi argued as follows:

The problem that I faced very early on was that such a structure would undoubtedly be expensive. The likelihood of London being flooded in a decade, or even in two decades, was not all that high, but I soon became utterly convinced that this must not be allowed to happen at all. Of course, following 1953, considerable safety precautions on the evacuation side were taken, and had they all worked I do not think there need have been any human casualties; but if you can imagine London with all the tunnels of the underground buried deep in mud and the whole tube system having to be effectively rebuilt, with the whole telephone system out of action, with much of the sewer system out of action because of mud in it, then you will realize the size of the catastrophe that a flood would have been. And it seemed to me clear that if it was possible to avoid such a thing, it had to be avoided; it was no use multiplying the number of pounds it would cost to clear up after the mess, with the probability of it occurring, to find a justifiable value of the investment. True, had the cost been fifty times what it turned out to be, one might have said no. But, given that the cost, though high, was perfectly affordable, and that the probability of the disaster occurring was significant, it seemed to me that its exact value was, for all practical purposes irrelevant.

If there had been a scientist in Israel who could have persuaded the Israeli government in 1972 to adopt this policy in relation to railway intersections the ghastly tragedy at the Habonim junction some ten years later, in which a bus was hit by a train and 50 children on a trip lost their lives, would almost certainly have been averted.

3. Facts and Figures
There are four major sources of accidents in Israel, traffic, industry, home and leisure, and natural hazards. Let us examine the statistics of each.

About 500 people are killed on the roads each year, but for each person killed 9 are seriously injured, and 50 lightly injured. Seriously injured means suffering some incapacity for life; it may mean being confined to a wheelchair, or having sight or hearing permanently impaired, or suffering irreversible brain damage. Lightly injured includes hospitalization for weeks or even months, but with an eventual total cure.

It is reasonable to regard traffic accidents as the worst category because of high numbers and their severity, and because the bulk of those killed and injured are in the prime of life.

In 1992, 170 people were killed, 3500 seriously injured and 75,000 lightly injured in industrial accidents. The estimated cost to industry lost days of work is 900 million shekels; this is in addition to the cost to National Insurance.

Regrettably there is little statistical information available on home and leisure accidents. There were about 200 fatalities in 1989 due to falls, and a total of about 700 fatal accidents other than traffic and falls. This miscellaneous category is supposed to include home accidents like electrocution and poisoning, leisure accidents like drowning, helicopter or plane crashes, deaths from natural hazards like flooding or lightning; but no detailed breakdown is available.

4. Road Accidents
Very detailed statistics are published by the Israeli Government Office each year. But before drawing conclusions one must think carefully about the significance of the different statistics. If we are interested in the social impact of accidents we should study the total number of killed and injured, and relate it to the number of residents. Table 1 taken from a publication of the Ministry of Transport shows the figures for the period 1982-1992.

Table 1: Road Accident Casualties During the Decade 1982/1992

Holland U.K. Switzerland U.S.A West Germany Belgium France Italy Austria Israel Ireland

3.3 3.7 4.8 6.5 6.9 9.5 11.0 12.3 13.6 21.1 38.3

(Cited in the Livneh Committee report)

Road accidents arise from a variety of causes each of which merits detailed analysis, but the most important in Israel today are excessive speeding, and failure to maintain an adequate distance between successive vehicles (tail-gating). Elementary physics tells us that the destructive power in an accident, the kinetic energy, increases as the square of the speed. But there are other speed dependent factors like the distance traveled before coming to rest which is governed by the reaction time and initial speed. It is an empirically established result that the fatality rate depends on the fourth power of the average speed. This has been widely tested both in the case of increase and decrease of the average speed. It means, for example, that an increase of average speed from 90 kph to 100 kph will be accompanied by an increase of 52% in fatalities, quite a startling result.

At this point I should like to correct a popular illusion that increased traffic density and the resulting traffic jams lead to increased fatalities. It has been pointed out many times by Dr. Elihu Richter, one of the leading Israeli researchers in the field, that just the opposite is the case. Traffic jams cause frustration, and possibly even nervous breakdowns, but not fatalities. The ultimate limit of a traffic jam is a state with all vehicles at rest; no one gets anywhere, but no one is killed or injured. Empty roads late at night or early in the morning, which create the illusion that there is no one else on the road, and invite excessive speeding, are a regular source of fatalities.

The choice of speed limit is determined as a compromise between the desire to move quickly from place to place and the need to avoid fatalities and injuries. Until 18 months ago it had been, since 1977, 90 kph on fast inter-urban roads in Israel. Little effort was made to enforce this limit, and anyone who has driven a car at 90 kph on the Jerusalem-Tel-Aviv road will testify that he is overtaken by 90% of the cars on the road. A steady state was established with an average speed of perhaps 105-110 kph.

There are strong vested interests lobbying for this limit to be increased. Past Ministers of Transport have resisted the pressure with the valid argument that they did not wish to add to the severe accident toll. The present Minister of Transport deviated from this policy, and seems to have been persuaded that the new Israeli road building plans necessitated an increase in the speed limit. In any case he appointed a committee of experts under the chairmanship of Prof. Moshe Livneh of the Technion to examine the general question of speed limits on different types of road, and to make appropriate recommendations. There were 8 members of the committee with representatives of the Ministry of Transport, the Police, the council for the Prevention of Accidents, the Road Building Authority, as well as scientists from the Technion. The report of the Livneh Committee is comprehensive and well documented. Its recommendations can be summarized as follows:

1. Within cities where the normal speed limit is 50 kph special zones should be created near schools and other places of congregation where the limit is reduced to 30 kph.
2. On well constructed fast inter-urban roads the speed limit should normally be 110 kph, but reduced to 100 kph or lower where the situation requires it.
3. For recommendation 2 to be adopted it is essential that a massive effort be made to establish a proper enforcement procedure.

Recommendation 1 above could represent a life-saving step forward. It has been successful in other countries in Europe; for example in Kingston on Thames, near London, the number of accidents has dropped to zero since it was introduced 2 years ago. No action was taken by the Minister of Transport in relation to this recommendation.

But in response to recommendation 2 the minister decided to initiate an "experiment" on two selected stretches of road on the Tel Aviv-Jerusalem and the Tel Aviv-Haifa highways, where the limit was raised to 100 kph on November 1, 1993. Little was done to implement recommendation 3.

In the 12 months which followed this action the number of killed on Israeli roads increased by 12%; there were 60 more fatal casualties than in the corresponding period of the previous year. Such an increase cannot be attributed to statistical fluctuations, and merits a scientific analysis. It should first be pointed out that because of the fourth power law an increase of only 3% in average road speeds is sufficient to account for the increase in casualties. Secondly it is unrealistic to think that the influence of an increased speed limit can be confined to two stretches of road; there is a well known spill over effect which extends the relaxation to other road users. The argument used by supporters of the change in speed limit that no significant increase has been recorded on the two designated stretches of road has no statistical validity; in dealing with a Poisson distribution of rare events a small sample is not representative. Further, the hope expressed in the Livneh report that since most drivers already exceed 100 kph there might be no further increase as a result of the relaxation is at variance with previous experience of human behavior.

One member of the committee, Dr. Dan link of the Ministry of Transport, was perceptive enough to anticipate the probable consequences of a recommendation to increase the speed limit, and he had the courage to differ from his colleagues and his boss, and oppose recommendation 2.

Recommendation 3 must surely command the support of anyone who has thought about the problem. Israeli technology which would enable enforcement to be effective has been available commercially for 5 years, but the Ministry of Transport and the Police have failed to exploit it properly. It was initiated by a former president of this Society, Prof. Gerry Ben David, who started life as a particle physicist, and decided to devote time and attention to the road accident problem when several of his friends were killed in a road crash. Ben David felt that the two essential requirements for progress were rapid and accurate measurements of speeding and tail-gating, and modern data processing which could handle the mass of information which would be required. He proposed a system consisting of two infra-red beams which would be directed across a traffic lane, and, by electro-optical measurements of time, the speed of all vehicles crossing the lane, and the distance between successive vehicles could be determined. A camera would be included in the unit, and computer control would arrange to photograph the registration numbers of cars exceeding a particular speed, or failing to maintain adequate separation.

Such a system, called the Marom, was developed by Prof. Joseph Bodenheimer and his colleagues at the Jerusalem College of Technology, and its performance was up to expectation. The commercial production was sponsored by a British philanthropist, Mr. Sidney Corob, who had been deeply disturbed by the road accident situation in Israel.

The Marom can provide 50 times as much data as a radar system. It can readily moved from place to place and can be on duty 24 hours a day; this is particularly important for identification of those who speed on empty roads at night. It is to be expected that drivers will learn to spot Maroms, and might slow down only in their neighborhood, and speed up again when they have left the area; this can be countered very effectively by stationing two Maroms, one at each stretch of road and coordinating their data. For example, on the Arava road, where there is little opportunity to turn off, a Marom could be stationed at each end of the road, and a driver who speeded up after leaving the first Marom would have a shock to find at the exit that his average speed over the whole journey had been measured.

By stationing a number of Maroms on Israeli roads motorists would soon gain the impression that they were under continuous surveillance. The manpower requirement to achieve this aim is very modest, and far less than increases in traffic police which have been envisaged using current conventional equipment. Of course it would be a type of "big brother" situation, but surely allowable as a life saving measure.

Unfortunately there has been no systematic use for the Marom so far on inter-urban roads. An urban programme was carried out by Ben David last year in Rehovot which succeeded in cutting down average speeds, and reducing the accident rate by 15%; during the same period the accident rate increased in all other major cities in Israel.

A number of features of Ben David's programme merit attention. Offenders are stopped on the spot; this is achieved by stationing a second unit a few hundred metres ahead of the Marom to which information is passed. Most offenders are in the category of 10-20 kph above the speed limit, and no fines are imposed for a first offense; instead an explanation is given of the aim of the project, and their co-operation is sought. As regards the actual change in driving behaviour, Ben David noted that when the project first starts up there is little interaction between drivers; the offenders resume their journeys and hopefully modify their own behaviour, but the general driving pattern is unaffected. But there is a critical stage at which the news of the Marom gets around, and the number of offenders drops rapidly. Any programme should be operated at least as far as this critical stage.

The results of the Rehovot project have been sufficiently encouraging for other cities (Petach Tikva and Netanya) to agree to run similar programmes.

5. Industrial Safety

It is gratifying to record that the government office which looks after industrial safety in Israel is run competently and efficiently. The Institute of Safety and Hygiene, directed by Dr. Menachem Schwartz, is a branch of the Ministry of Labour funded by the National Insurance Institute. It organizes educational courses of different types and lengths, and makes use of the latest computer equipment in its teaching programme. It arranges an Annual Conference and Exhibition with facilities for the display of different types of safety equipment. An information centre is maintained in which the knowledge and experience of the Western world on industrial safety are stored systematically using the latest modern technology. The managers or workers of industrial plants can telephone for advice on any problem which they face, and can be assured of a rapid and reliable response.

A regular participant in the educational programme is Prof. Trevor Kletz, a top international authority on industrial safety who is a lucid and experienced lecturer. Kletz, who was trained as a chemist, spent many years working at ICI in production management, and then became safety advisor to the Petrochemicals division. He has a distinguished academic record, and was elected to a Fellowship of The Royal Academy of Engineering in 1984; he holds a visiting appointment in the Chemical Engineering Department at Loughborough University.

I recommend all of Kletz's books, but in particular a recent publication entitled Critical Aspects of Safety and Loss Prevention (Butterworth 1990) which is a collection of nearly 400 pertinent observations on safety. The title was chosen in relation to his professional colleagues, but from the point of view of the non-specialist a more appropriate title would be A Manual of Common Sense and Safety. Here is a characteristic sample:

Human failing
In some companies 80-90% of the accidents that occur are said to be due to human failing, that is, a failure by the injured person or fellow co-workers rather than a failure by the manager, supervisor, or designer. . . .. In fact, as I shall try to show, most accidents can be prevented by better management or better design. . . .

To say that accidents are due to human failing or human error is not so much untrue as unhelpful. It does not lead to effective action, only to advice to take more care. Instead of asking what is the cause of an accident we should ask what we can do to prevent it happening again. We may then think of ways of improving the training, supervision, design, and so on.

Kletz is a pioneer of the inherently safer design approach, and always emphasizes that this should take place at the initial planning stage, and not after an accident has occurred. He was one of those responsible for introducing HAZOP studies as a mechanism for anticipating, and hence avoiding, potential accidents. This is his summary of the method:

Hazard and Operability Study (Hazop)
A hazard and operability study (hazop) is the preferred method, in the process industries, of identifying hazards on new or existing plants. At one time we identified hazards by waiting until an accident had occurred; we then took action to prevent it happening again. This, 'every dog is allowed one bite' philosophy was acceptable when plants were small, but is no longer viable now that we keep dogs as big as Flixborough or Bhopal. We need to identify hazards before accidents occur. Check lists have the disadvantage that new hazards, not on the list, may be overlooked so we prefer the more open-ended hazop technique. It allows a team of people, familiar with the design, to let their minds go free and think of all the deviations that might occur but it is done in a systematic way in order to reduce the chance of missing something.

Let me remind you that the Flixborough chemical plant in England, which oxidized cyclohexane, was destroyed in 1974 by and explosion which killed 28 men on site and caused extensive damage and injuries in the surrounding villages. Bhopal, a town in central India, was in 1984 the scene of the worst disaster in the history of the chemical industry. A leak of a toxic chemical spread beyond the plant boundary, killing about 2,000 people and injuring about 200,000. Kletz points out that a key contribution to both disasters was the unnecessary storage of large quantities of hazardous materials.

In spite of the laudable activities of the Institute of Safety and Hygiene, it is distressing to note that Israel has the worst record among developed countries of the number of accidents relative to the size of the working force. This statistic was revealed by Yossi Tamir, Director of the National Insurance Institute at the lst Annual Conference of the Institute for Safety and Hygiene. The vital human and economic importance of industrial safety does not seem to have penetrated yet to many Israeli industrialists.

6. Home Safety
We read regularly in the papers of fatalities at home arising from fire or smoke inhalation, of children choking or poisoning themselves, of deaths by electrocution or inhalation of gas, of drownings at sea or in rivers or lakes. From the statistics given in section 3 it is clear that the total number of fatalities other than those occurring on the roads or in industry is about 700 per year, but a detailed breakdown is of major importance if this toll is to be reduced. No one seems to have specific responsibility for education in this area, and one of the aims of the Nebenzahl Institute is to fill the void.

7. Natural Disasters
This is the conventional description of deaths, injuries, and economic losses arising from floods, droughts, earthquakes, storms, volcano eruptions, landslides, etc. But it conveys the unfortunate impression that nature is responsible for the disaster. At an International Conference on this topic which I attended in London in October 1993 several of the leading speakers pointed out that there are natural hazards whose origin and character are well understood by now. The disaster arises as a result of our failure to cope with these hazards. Kletz again has something illuminating to say on this topic:

Acts of God
This phrase is used to describe natural phenomena such as lightning, earthquakes and floods. It is an unfortunate expression which we should avoid, as it implies that we can do nothing about them. However, to quote from a book on natural disasters (A. Wijkman and L. Timberlake, Natural Disasters -- Acts of God or Acts of Man?, International Institute for Research and Development, London, 1984, pp. 6, 29 and 30),

Some disasters (flood, drought, famine) are caused more by environmental and resource mismanagement than by too much or too little rainfall. The impact of other disasters, which are triggered by acts of nature (earthquake, volcano, hurricane) are magnified by unwise human actions. Humans can make land prone to flooding by removing the trees and other vegetation which absorb this water.

Humans can make land more drought-prone by removing the vegetation and soil systems which absorb and store water . . .

In other disasters such as cyclones and tsunamis, humans can increase their vulnerability by destroying bits of their natural environment which may act as buffers to these extreme natural forces. Such acts include destroying reefs, cutting mangrove forest and clearing inland forests.

The Ethiopian famines of the 1980s were due to mismanagement, not drought. Israel, a food exporting country, has a lower rainfall than Ethiopia but in that country over-grazing and deforestation have resulted in a loss of topsoil and there is little irrigation.

Similarly, suppose a storage tank containing petrol or other flammable liquid has been blown up by lightning. We cannot prevent the lightning but we can prevent the results by using a floating-roof tank or nitrogen blanketing or installing a flame trap in the vent line and by seeing that this protective equipment is kept in working order. The explosion could have been prevented by better design and/or better methods of working.

The uses of phrases such as 'cruel fate' implies helplessness and inevitability and discourages a search for ways of preventing the accident or disaster.

National scientific academies have begun to interest themselves in reducing the toll of natural disasters, and have combined forces to launch a programme of activities under the heading IDNDR (International Decade for National Disaster Reduction). A leading proponent of this programme is the distinguished physicist Dr. Frank Press, former President of the US National Academy of Sciences, who in a lecture delivered to the Royal Society stated the following:

There is much that is known about disaster-proof design and construction, which if disseminated could reduce the severity of loss. In many cases prediction of catastrophic events can save many people. Improved public policies relating to land use, construction codes and standards, and education can contribute to disaster mitigation. In all of these there is an extraordinary opportunity for scientists and engineers to make a valuable contribution to public welfare.

This thesis is well borne out by the following recent earthquake statistics. In October 1993 an earthquake of magnitude 6.5 on the Richter scale struck a village in India and claimed 10,000 fatalities. A year later a more severe earthquake struck in California; the papers were full of stories of harrowing experiences, of people trapped in elevators whose shafts were totally distorted etc., but the number killed was less than a hundred.

Two natural hazards should be of major concern to Israel, droughts and earthquakes. This year has witnessed an exceptionally good rainfall, but cycles of good and bad years are a feature of the rainfall of our region. Benoit Mandelbrot has even written a paper in Water Resources Research 4,909 [1968] entitled "Noah, Joseph, and Operational Hydrology" in which he shows that "The Joseph Effect" and "The Noah Effect" are non-Gaussian, and he proposes alternative methods of fitting based on self-similarity. But unlike Joseph we make little effort to store the bounty of the good years, instead we open the sluice gates at Degania and raise the level of the Dead Sea. In the bad years we have talked about importing water from Turkey.

The ancients devised sophisticated methods of water storage which enabled them to get by. Their requirements were very much more modest than ours, but I am convinced that, if we devoted serious attention to the problem and harvested modern technology, we could meet our own requirements.

Anyone familiar with the T'nach knows of its references to earthquakes; a particularly severe example which occurred during the reign of King Uzziahu was used for some time afterwards as a date reference (Amos 1:1, Zechariah 14:5). The excavations in Bet Shean show that the city was totally destroyed by an earthquake in Roman times. More recently an earthquake in 1837 destroyed the city of Safed and claimed 5,000 victims; an earthquake north of Jericho in 1927 claimed hundreds of victims over a wide area.

Past records indicate that a severe earthquake occurs about once per century in our region and this is a clear example of a rare and horrendous event which I discussed at the beginning of my lecture. Officially all building and construction work in Israel has to conform to earthquake-resistant standards. But it is surely time for these standards to be reviewed in the light of recent world-wide experience, and for an assessment to be undertaken of the extent to which the standards are enforced.

8. Conclusions
I have two recommendations to make in conclusion. Firstly, that safety education should be given a place in the academic curriculum, and that every University student should be made aware of elementary facts of the type that I have put before you.

Secondly that our top scientific manpower should be mobilized to help tackle the problems which I have outlined. I started my career in radar research for the British Navy during World War II, and was deeply impressed by the impact which first rate scientists could make in a new field when they were convinced that it was important. Experience at Los Alamos in the USA confirms this view. It is no exaggeration to say that what was then achieved in 3 or 4 years would have taken 20 years or more with regular professional personnel. I have described now Bondi was mobilized by the British government to advise on an important issue: Richard Feynmann and John vonNeumann were used similarly by the US government. The government of Margaret Thatcher initiated a think tank headed by Lord Rothschild to serve as a permanent advisory unit.

Israel's greatest asset is her reservoir of scientific brainpower which has been so well reinforced by Russian immigration.

Establishing think tanks at the top level could lead to real progress in the many problems which the country faces including those to which I have referred.

May I finally conclude with the observation that my efforts in this area will have borne fruit if I have succeeded in interesting a number of physicists in the topics which I have discussed.