Accident Analysis - Science topic

Dear All,

Just a month until the deadline of this call!

I am looking forward to receiving your articles.

Best regards,

FM.

Have you thought about using Stan or brms?

1.Accidents, Traffic - statistics and numerical data. 2.Accidents, Traffic - trends. 3.Wounds and injuries - epidemiology. 4.Safety - statistics and numerical data. 5.Data collection. I.World Health Organization. Dept. of Violence and Injury Prevention.

1 Social Impact Appraisal. 2 Accidents Impacts. 3 Physical Activity Impacts. 4 Security Impacts. 5 Severance Impacts. 6 Journey Quality Impacts and etc.

A Joint Report by the United Nations Office on Drugs and Crime and the Latin America and the Caribbean Region of the World Bank

UNITED NATIONS DEPARTMENT OF ECONOMIC AND SOCIAL AFFAIRS

Commission on Sustainable Development. Nineteenth Session. 2-13 May 2011

Dear Dr Shah,

I have seen in the thermal-hydraulics modeling of SBOs, this modification (injection of some numbersof accumulators into the SGs instead of primary circuit) will result in increase of coping time, but I think the answer to your question is very dependent to the NPP design and also consideration of other accidents. In other words it should be assessed case by case.

First, I consider an under operation PWR. This Plant is designed with specific numbers and configuration of accumulators, HPIS, LPIS and the associated accident analysis including (LOCA, SBO, MSLB, and ...) is performed considering these safety injections. Modification of Accumulators injection lines needs Re-assessment of all theses accidents and performance analysis of other injection systems, as the modification may be useful for SBO stand alone, but it may deteriorate the situation for some break sizes of LOCA or other accidents. Furthurmore, backfitting of the NPP should be assessed from the structural-civil work aspects and also some new setpoints, as this kind of modifications may need closing/ opening new penetration in the SG's piping, compartments of the containment and etc. but it doesn't mean that it's not applicable, this modification is more practical for NNP designs with larger number of accumulators or other sources of water that is capable of injecting water into SGs. You can find some studies that the SG shell side can be refilled by use of fire fighting hoses that are fed from outside of the containment.

For newer designs of NPPs and under construction NPPs, I have done some literature search and what I found is that the designers are more tended to use passive heat removal systems for condensing steam from SGs and send it back to them. these kind of systems can extend the coping time much more in comparison to accumulators injection into SGs.

Best Regards.

Alireza

Just an update - the current IF of Applied Sciences is now 2.217 (was 1.689).

Deadline for manuscript submissions: 31 July 2019

Some motorists say cyclists are stealing their space in the road and that they shouldnt be allowed to use the roads.

In a personal experience, I was told I was "slowing" them.

I think bad experience with irresponsable cyclists could influence, in mi city there are a lot of cyclists who ignore red lights, make dangerous turns, etc, so some motorists use this as an argument.

The basic idea is to detect changes in pitch, roll, and yaw by the sensor.

You can use the Kalman Filter for the real-time prediction.

You can find some statistics when scan data base by age category: www.sewik.pl - polish database of accidents.

The linear correlation between motorcycle crash speed, crash speed squared (as a measure of KE) and various measures of injury severity (ISS, AISmax, Brain AISmax, etc.) runs in the neighborhood of .25 - .35.  This was true for Hurt study, Thailand study and MAIDS data.   Terry Smith's 2009 MAIDS paper showed a significant correlation of precrash (running) speed and injury severity but not crash speed.  

Agree with the view that going from injury severity to impact speed would be too reliable. Having done a lot of research review in this field in the past four years, I am not aware of any published work on this.

Determining delta-v from vehicle crush depth, skid marks and other crash dynamics evidence is a pretty standard crash reconstruction practice. There are publicly available US data sets of reconstructed crashes (+200,000 cases) held by National Highway Traffic Safety Administration (NHTSA). The data sets are called NASS/CDS and CIREN databases. Injury data is available there as well. Cases are weighted according to distribution of crash severities in the USA.

Delta-v can be used to theorise vehicle impact speeds, and the relationship to injury severity, although many limitations exist. We did this in a couple of recent papers among my uploaded publications.

It is worth checking the experience gained in large naturalistic driving studies (NDS), for example SHRP2 (https://insight.shrp2nds.us/documents/shrp2_background.pdf). The issue is not how to collect data but how to process it afterwards. One has to define and apply various (semi)automatic triggers and also link them to the environment characteristics.

A recent Austroads study authored by my colleague at ARRB:

HI, if this is still an active question for you try the excellent answers and references I received for my recent question on urban local road safety. Most of the answers related to urban road safety in general.  

Very roughly: intersections and their design, land use, road standards, pedestrian movements.

If you can not contact the authors contact me and maybe I can help.

Look at the Nationalk Highway Traffic Safety Administration's regulatory impact analysis on tire pressure monitoring systems at  http://www.nhtsa.gov/DOT/NHTSA/Rulemaking/Rules/Associated%20Files/TPMS-2005-FMVSS-No138.pdf

Especially look at the analysis of impact on crashes.

BRSI link to weather and road safety research appears to have been fixed:

According to my personal experience just because I worked in the past as a journalist for a specific campaign about road safety, I studied a bit the subject.

The client was a very relevant Italian PA with a consistent budget.

We share with those guys a panel of ideas and then they developed  a specific capaign utilizing an inhouse company.

The final result was to differentiate  the campaign through different medias .

They chose 3 different channel because it was in the late '90 during wich the social media had practically no existence.

So we proposed to select TV, radio and newspaper (divided in two groups, ie generalizated and specialized. The difference was to considered specialized if dedicated to a target focus of one scientific discipline.  So, for example, a financial newspaper was considered specialized instead a monthly journal for describing new and old cars was considered generalized).

In TV and in the radio the choice was to advertise in some particular hours like 6.00- 8.00 am and 6.00- 9.00 pm.

The message of 30'' or 1' was to describe what happened before an accident with deaths in wich, e. g., a child ask to his father: "dad, why did you decide to use your mobile phonewhen I say do not? " or a wife taking with her husband " you were in late but I didn't expect for dinner on time. I knew that were snowing. I told your on the phoneymoon be fore your leave the work". Others were about the the abuse of alcohol or drugs and even the cigarettes (the last was stopped by some tobacco lobby).

For specialized newspapers  we proposed to show inphograph about stats of accidents, number of deaths and serious  injuries with historic trends ( hoping they showed an improvement)

 For generalizated newspaper the message was with pictograms showing images of serious casualties and accidents.

Hoping to be a little bit useful. 

Luca

You may want to read the following paper. It is about application of WAD of pedestrians and bicyclists for head injuries. 

Option 1 would a validation exercise indeed. You can do this very simply by correlating the model and real outcome. There are several statistical methods depending on what outcomes you are comparing. One example I have is EuroRAP star rating vs. crash rates - two different variables, not directly comparable.

If you went with Option 2, e.g. a statistical modeling (accident prediction model), then you can use various residual plots as long as you got an independent sample not used in modelling. There are more sophisticated methods than that of course.

Option 3 could be modified by adding data from 'safe' sites. Then you are looking cross-tab analysis, the predecessor of statistical models. See answer for Option 2.

Option 3 - you are looking at proxy measures. As far as I understand this subject, it is the next frontier in road safety. Great for developing understanding of combined human, vehicle and road factors involved in crashes, i.e. everything that current statistical models do not do. The near future.

Hello Lin Zhou

please check pdf

To add to this: in this report

the scientist involved have found that in the age group 20 - 60 years old, male drivers had 1.5 higher risk (accidents/km) of a fatal accident than female drivers. However, for the non-fatal accidents, the risk of the women is 26% higher to get injured and 16% higher to get involved in an accident than men. So, one may in fact assign males a more risky driving-style, but only after proper normalization. However, this analysis was performed for the US in 1990; different areas and different times may provide different numbers.

Hi all,

I have published a new paper on this area, which uses mental mapping and a Generalised Linear Mixed Model to unpack perceived cycling risk: http://www.sciencedirect.com/science/article/pii/S0001457515301688

Richard

Yes, congestions also came up my mind... But since AADT is usually used (which is aggregated annually), I would think it is not likely that congestion would influence this.

You are correct, both AADT and length are the typical exposure variables on a segment level.

One possible source of information - http://safety.fhwa.dot.gov/rsdp/downloads/spf_development_guide_final.pdf

 Thanks Deborah. Looks really useful. I'll read the paper. Is the ontology open access?

Mark, Appreciate the offer. I'll keep it in mind. Thanks.

Crash Severity on a urban highway, as on other roads, mainly depends on the kinetic energy. The kinetic energy by the speed. The speed by traffic flow, density, lane width, tortuosity of the route etc.

I would check the following parameters in function of the crash  severity:

number and width of lanes, traffic flows per lane (in passenger car equivalents per hour), composition of traffic flow, speed (85th percentile or speed diagrams for the entire path), speed differences among different types of vehicles (such as cars, vans, trucks etc).

The interdependence of these parameters and the crash severity should give you some good results.

Good work and good luck

In Europe, speed limiters are mandatory, and there are specific speed limits that the drivers must comply with, according to the type of roads. The itinerary must be provided to the law enforcement.

All heavy vehicles (for passengers and goods) must be equiped with a tachograph (analoge or digital) wich allows the law enforcement authorities to track the driving and rest periods, as well as the speed.

European laws are quite strict when it comes to avoid fatigue in driving so resting periods are compulsory and the company must keep the record of the data for control purposes. The driver must be able to provide the authorities the data of the last 28 days so that he can prove that he hasn´t been working more than he should.

The fines are heavy for the companies but, as in almost everything, the companies still find a way to get around the rule of law.

Dear Quan Yan,

Yes, you can continuously run the simulation post-crash, depending on your  simulation engine/software. We developed our own for research purposes and thus we are able to control the behaviour of the simulation software like what we want. Perhaps you might want to contact simulation professional software developers ( like drive square above ) for assistance. 

The European Council of 24/25 March, 2011, requested that the safety of all EU nuclear plants should be reviewed, on the basis of a comprehensive and transparent risk and safety assessment ("stress tests"). These “stress tests” are defined as targeted reassessments of the safety margins of nuclear power plants, developed by ENSREG, including the European Commission:

All nuclear power plants in the EU underwent stress tests and peer reviews in 2011 and 2012. Many other countries and territories also conducted comprehensive nuclear risk and safety assessments, based on the EU stress-test model. These include Switzerland and Ukraine (both of which fully participated in the EU stress tests), Armenia, Turkey, Russia, Taiwan, Japan, South Korea, South Africa and Brazil:

Follow-up of the peer review of the stress tests performed on European nuclear power plants - see:  http://www.ensreg.eu/EU-Stress-Tests/Follow-up

The March 2011 accident at the Fukushima Dai-ichi plant resulted in a new effort on the part of the nuclear industry to ensure that the existing SAMGs were as effective as possible. EPRI Report „Severe Accident Management Guidance Technical Basis Report Volume 1: Candidate High-Level Actions and Their Effects“, Final Report, October 2012, updates the original technical bases for SAMGs to reflect the lessons learned to date from the Fukushima Dai-ichi accident:

This update also incorporates insights gained from a substantial body of severe-accident research and analysis conducted since the original development.

you must use new technologies  like Intelligent transportation systems 

If you have no denominator data for the sites (eg no of vehicles passing/unit time, population density of area) you could use a comparison of Poisson variables ,using  crash counts in successive time periods as basic data.

If you have denominator data you could use comparison of proportions with exact binomial statistics or normal approximation.

When discussing correlations you have to keep in mind that correlations do not necessarily imply causal relations. Often, they are a consequence of a common influence by an ignored third factor such as properties of the infrastructure (many or few freeways), regional structure (predominantly urban or countryside), cultural properties,  (liberal or strict) legislation, (sloppy or stringent) control/enforcement. For example, in countries with liberal speed regulations (e.g., 75 or even 85 mph allowed on some roads) and sloppy enforcement (hardly any speed limit monitoring), few speed tickets are issued, i.e., the statistical data show little speeding. Nevertheless, in an objective sense, there may be many drivers driving too fast for the given situation resulting in comparatively many accidents.  This results in a paradoxical correlation with an unexpected sign between speeding and accidents.  This is a variant of the classical Simpson's paradox: The more firefighters are sent to the scene, the more damage the fire will  cause.

We are doing some research on prevalence of injuries among players Chhattisgarh, India.

Dear Thais, 

Random-effects models become necessary when spatial correlation (data collected from the same geographic region) or temporal correlations (data collected con the same road section over successive time periods) are suspected to be present. It is to be said that random-effects  models are more especially appropriate for capturing respectively spatial or temporal correlations, but not a combination of the two.

Random-parameters models attempt to account for possible unobserved heterogeneity (unobserved factors that may vary across accident observations) by allowing the regression model coefficients to be random-parameters, in contrast with the assumption that parameters are considered to be constant across observations.

A special case in the class of random-parameters models is the so called random-intercept Negative Binomial model. In this model only the regression intercept is random, and this model is known in the literature to be equivalent to the random-effect Negative Binomial model.    

A good paper that discuses this issue is "How Many Accidents Are Needed To Show A Difference?" by Ezra Hauer.  Here is a link to the article: http://www.sciencedirect.com/science/article/pii/S0001457508000535

@Xochitl

Experience can certainly be a factor, but one which is hard to disentangle from a general decrease in risk taking behaviour with age (perhaps based on near misses or real accidents)

@Vincenzo

Distraction is certainly one of the top causes cited by cyclists involved in an accident in Belgium. However is it unclear to me why women would be more distracted than men (in Belgium and the other way around in other countries)

@Yves 

The population was limited to those of working age and mainly focussed on commuting so probably not due to differences in time activity pattern.

However on the gender issues I think you may have a point. Some of the men in the study reported really long commuting distances along e.g. canals where accident risk (per km cycled) is probably very low. On the other hand women seem to take trips more often in the rush hour (e.g. AB modelling by Beckx et al) and in town centres with more intersections and therefore higher risk (while their risk aversion while driving/cycling) may still be higher than that of men)

@Henk

Thank you for the Dutch graph. The reversal of male/female risk ratio with age is really interesting. The SHAPES sample was limited to those of working age (<65) and the number of observations in the +50y was indeed very low. On the other hand we are very confident in our assessment of the distance cycled which we observed frequently during a whole year.

In Brazil, the quality of service on rural highways are determined based on HCM. However, there are studies that have been produced to create appropriate methods to evaluate these roads. Some papers were published in english, but others are not published yet.

A good starting point is the HSE information sheet on Explosion hazards due to spray releases: http://www.hse.gov.uk/offshore/infosheets/is1-2009.pdf 

Dear Muath can you send me a copy of these paper?

Fines is the most effective method to improve road safety these days in my country.

This is a known dilemma. If you make the road safer, the driver takes more risks. We call ourselfs an "intelligent life form" but we don't really act very intelligent. Espacially not while driving.

I personally think that the roads will only then be really safe, when WE stop driving and let the computers drive our vehicles. But that is still a distant future.

In the meantime we try to compansate the psychological errors of drivers by environmental measures.

Another suggestion is to detect the deacceleration rate. for example if the object goes from high speed to zero speed in a short period of time, that would be an indication of accident.

Combined by analyzing the change of the speed in the surrounding objects, sudden changes of some objects would be an indication also.

fuzzy data classification methods are suitable for complex data subject to uncertainties. The uncertain nature of noisy data requires the use of fuzzy methods.

Good luck

Dear Sara, dear Niels,

the following thesis

says that an attempt to translate ICD-10 to AIS 2005 update 2008 was only partly successful. In general, for me it seemed unsatisfactory and not recommendable without a lot of future improvements. For ICD-9 I cannot tell, but I would not suggest to use AIS 90 or 98 today, as the 2005 version is now the most updated scale. A translation from 98 to 2005 is provided on

but still no perfect fit is given.

Anyway: as different approaches towards injury coding are realised in ICD and AIS a translation will always be very limited.

If you have access to raw data (injuries) I'd suggest to code directly in AIS. Of course it needs some time, but the quality will be much better.

Kind regards

Sylvia

You could take a look at http://www.bmj.com/content/328/7444/857 a study of motorcyclists in New Zealand. Beware though, they went to great lengths to trace and contact MC drivers at 4 recieving hospitals and got a good response rate. If you don;t have resources for this your rep[sonse rate will be much lower and probably quite biased.

Best of luck

Phil