10。It also depends on the experiences and attitudes of the team (Ramasesh and Browning, 2014), and the level of the vulner- ability and exposure of people (Cardona et al。, 2012)。 For example, a complicated hazardous situation may be per- fectly clear to the specialists working on it, but simply entraps a manager visiting the site。

11。If they are going to facilitate the process of risk identifica- tion, drivers must be measurable。 No doubt attributes like complexity, novelty or persity will increase either proba- bility or consequences of a risk, but no mechanism for track- ing such drivers is defined, especially when this is to be done automatically。

12。Drivers must be detectable by the tools in use。 Basic blocks in a BIM platform are building elements with all their rela- tions and attributes (Motawa and Almarshad, 2013),  so, some safety provisions, e。g。, the employer’s failure to provide safety instructions as required by OSHA 1926。21。b, cannot be considered as a risk driver when working with object ori- ented software,  although  they  may  considerably  influence a great amount of safety  issues。

4。Results

Application of Behm’s model on all 363 cases reveals that 46。8% of the accidents are related to the concept of Design for Safety。 Table 3 shows how this proportion changes with the source and type of accident records。 In this table, it is specified how many acci- dent cases can be related to design, in terms of the source narrating the accidents (OSHA, NIOSH, or the survey), and their type (Fall or Falling Objects, Unsafe Operation, Electric Shock, and Cave in or Asphyxiation)。 A general pattern of increased relation to DfS can be seen in the cases obtained from the survey, and also in the acci- dents categorized as Fall or Falling Objects and Cave in or Asphyx- iation。 For example, among the 27 fall incidents reported by OSHA fatal facts, 12 (44。4%) were related to DfS, and this proportion is higher by 36% in the accidents  reported  by  the  participants  of the safety course。

The items provided in this section express the safety risk drivers that are found playing important roles in the analyzed accident reports。 Each driver is considered to drive a certain risk category

by Solihin and Eastman (2015)。 For example, to check whether a working place is naturally lit, or to check if emergency medicals have access to the place where a victim has collapsed, both are pro- cessed by extended topological pathfinding calculations, and fall into class three。 Table 4 summarizes different types of problems that have to be solved during the rule checking process。 Each prob- lem type is labeled with a keyword and referred to one of the four rule classes。 It should be noted that some types of problems can be seen as inclusive of few other problem types, in which case the label of the more complicated problem type will be assigned。 As an instance, to solve a ‘‘Vicinity” problem, all the nearby objects have to be checked first to see if they can prevent the driver from affecting the target object, which results in an area  or  space defined as an effective region, and then, an ‘‘Existence” problem has to be solved over the part of model that lies in the effective region to find any object that can induce the driver。 Similarly, ‘‘Spotting” represents the procedure that examines all the objects existing in an effective region to see if any of them is able to pro- vide the sought-for functionality。 For example, to spot a potential anchorage point to which a worker might attach a safety harness, every hole like portion of metal  objects  have  to  be  assayed  for the strength of holding a worker’s weight, in the case that the lan- yard can reach the working  area。

Of a general concern are the circumstances explained in Table 5 that usually intensify the consequences of an accident。 A noticeable number of fatal cases were about injured workers who were left alone for a while because no one could be aware of the happening or gain access to the place。 Improper means of access to the work- places not only frustrated the efforts of rescue, but also actively ini- tiated some risks, as some untrained workers decided to create an easier path by removing barriers or using unsafe or unassigned tools。