**6. Subject matter experts**

The Subject Matter Experts (SMEs) have two objectives along the risk management study; firstly they provide specialized knowledge about specific project risks and uncertainties and


Project and Enterprise Risk Management at the California Department of Transportation 417

which is a very productive and healthy action for the project. The downside of conducting

In some cases, we can get different SMEs opinions for the same variable or risk as mentioned by Vose (2008). Experts will sometimes produce profoundly different probability distribution estimates of a parameter. This is usually because the experts have estimated different things, made differing assumptions or have different sets of information on which to base their

As can be observed, from the input contribution to the risk model from the SMEs is a combination distribution. This is not a common situation but there is always a chance for encountering these types of challenges. Especially for complex projects where a considerable expertise is required. Figure 4 illustrates an example of combining three differing opinions, but where expert A is given twice the emphasis of the owing to the greater experience of

It is relevant to notice that not all the SMEs are willing to participate actively in a risk management exercise at the first time, especially if they have no been exposed before into one. When SMEs are first asked to provide probabilistic estimates, they usually won't be

For overcoming this and in order to provide the SMEs some sort of support in the field of the probabilistic risk management, we held a sort of training and educational meeting with

It can be noted from Figure 5 that the expert's knowledge can extend beyond the knowledge base into the absolute truth area as a result of creativity and imagination of the expert.

opinion. However, occasionally two or more experts simply genuinely disagree.

these interviews is that they are rather time consuming.

**Figure 4.** Combining three dissimilar expert opinions (Vose, 2008)

particularly good at it because it is a new way of thinking (Vose, 2008).

that expert (Vose, 2008).

all the RMT and SMEs.

**Table 1.** Stakeholder analysis (Caltrans, 2007)

secondly; in lieu of having a set of data available, they can fill the gaps of insufficient data while defining the uncertainty of a project variable. In practice and with transportation projects, usually two or three variable values are asked to the SMEs. For other type of projects or industry this approach may be different for example if the project is very unique and complex, like an energy plant or even an oil platform.

If for example the goal is to work with two values for defining the uncertainty behaviour for a particular risk variable, then the probability distribution called Uniform Distribution is used, which assumes all the values between the minimum and the maximum values have the same probability of occurring. In this case, a worst and best case scenarios are asked from the SMEs for defining a minimum and a maximum value of the risk. The other common probability distribution function use in this field is the triangular distribution function. This function assumes that there are three different values assigned to a variable of risk; a minimum, a most likely and a maximum. It is difficult to define which function is better for obtaining the data from the SMEs, since it will depend of their experience, the type of project and the project data available at the moment of the study. In general, the uniform distribution is the most popular since it is rather easy to estimate only two values.

A key part for soliciting the information for the SMEs, are the questions asked by the Risk Manager or Risk Facilitator. In some cases, the SMEs are reluctant to participate and optional methods should be place on the table by the risk manager for getting the opinion needed from the SME. An alternative and perhaps a highly recommended approach is to have one on one interviews with the SMEs for asking the questions and getting the answers. However, the recommendation is to have this information during the meetings, so the rest of the team can discuss and in some cases even do they can challenge the SMEs opinion, which is a very productive and healthy action for the project. The downside of conducting these interviews is that they are rather time consuming.

In some cases, we can get different SMEs opinions for the same variable or risk as mentioned by Vose (2008). Experts will sometimes produce profoundly different probability distribution estimates of a parameter. This is usually because the experts have estimated different things, made differing assumptions or have different sets of information on which to base their opinion. However, occasionally two or more experts simply genuinely disagree.

**Figure 4.** Combining three dissimilar expert opinions (Vose, 2008)

416 Risk Management – Current Issues and Challenges

Project Sponsor

Project Manager

Project Manager Assistance

**Name Function Contact** 

Progress/Status

Progress/Status Meetings and Report. Information sharing and issue resolution

Progress/Status Meetings and Report. Information sharing and issue resolution

**Table 1.** Stakeholder analysis (Caltrans, 2007)

and complex, like an energy plant or even an oil platform.

Report

**Information**

**Preferred Method of Communication**

email email Time, cost,

email email Time, cost,

secondly; in lieu of having a set of data available, they can fill the gaps of insufficient data while defining the uncertainty of a project variable. In practice and with transportation projects, usually two or three variable values are asked to the SMEs. For other type of projects or industry this approach may be different for example if the project is very unique

If for example the goal is to work with two values for defining the uncertainty behaviour for a particular risk variable, then the probability distribution called Uniform Distribution is used, which assumes all the values between the minimum and the maximum values have the same probability of occurring. In this case, a worst and best case scenarios are asked from the SMEs for defining a minimum and a maximum value of the risk. The other common probability distribution function use in this field is the triangular distribution function. This function assumes that there are three different values assigned to a variable of risk; a minimum, a most likely and a maximum. It is difficult to define which function is better for obtaining the data from the SMEs, since it will depend of their experience, the type of project and the project data available at the moment of the study. In general, the uniform

distribution is the most popular since it is rather easy to estimate only two values.

A key part for soliciting the information for the SMEs, are the questions asked by the Risk Manager or Risk Facilitator. In some cases, the SMEs are reluctant to participate and optional methods should be place on the table by the risk manager for getting the opinion needed from the SME. An alternative and perhaps a highly recommended approach is to have one on one interviews with the SMEs for asking the questions and getting the answers. However, the recommendation is to have this information during the meetings, so the rest of the team can discuss and in some cases even do they can challenge the SMEs opinion,

**Goals on this project** 

quality

quality

email email Budget Monthly

**Frequency** 

Monthly

As occurs

As can be observed, from the input contribution to the risk model from the SMEs is a combination distribution. This is not a common situation but there is always a chance for encountering these types of challenges. Especially for complex projects where a considerable expertise is required. Figure 4 illustrates an example of combining three differing opinions, but where expert A is given twice the emphasis of the owing to the greater experience of that expert (Vose, 2008).

It is relevant to notice that not all the SMEs are willing to participate actively in a risk management exercise at the first time, especially if they have no been exposed before into one. When SMEs are first asked to provide probabilistic estimates, they usually won't be particularly good at it because it is a new way of thinking (Vose, 2008).

For overcoming this and in order to provide the SMEs some sort of support in the field of the probabilistic risk management, we held a sort of training and educational meeting with all the RMT and SMEs.

It can be noted from Figure 5 that the expert's knowledge can extend beyond the knowledge base into the absolute truth area as a result of creativity and imagination of the expert.

Therefore, the intersection of the expert's knowledge with the ignorance space outside the knowledge base can be viewed as a measure of creativity and imagination. Another expert (i.e., Expert B) would have her/his own ellipses that might overlap with the ellipses of Expert A, and might overlap with other region by varying magnitudes (Ayyub, 2000).

Project and Enterprise Risk Management at the California Department of Transportation 419

**Figure 6.** Overview of the Expert Elicitation Process (USEPA, 2011)

The data have simply never been collected in the past

The data are sparse, requiring expert opinion "to fill in the holes"

data and theory (Ayyub, 2011).

environment, etc)

The data are too expensive to obtain

The area being modelled is new

Expert Elicitation (EE) is a multidisciplinary process (Figure 6) that can inform decisions by characterizing uncertainty and filling data gaps where traditional scientific research is not feasible or data are not yet available. Although there are informal and nonprobabilistic EE methods for obtaining expert judgment. The goal of an EE is to characterize, to the degree possible, each expert's beliefs (typically expressed as probabilities) about relationships, quantities, events, or parameters of interest. The EE process uses expert knowledge, synthesized with experiences and judgments, to produce probabilities about their confidence in that knowledge. Experts derive judgments from the available body of evidence, including a wide range of data and information ranging from direct empirical evidence to theoretical insights. Even if direct empirical data were available on the item of interest, such measurements would not necessarily capture the full range of uncertainty. EE allows experts to use their scientific judgment transparently to interpret available empirical

As mentioned by Vose (2008), it is usually impossible to obtain data from which to determine the uncertainty of all the variables within the model accurately, for a number of reasons:

Past data are no longer relevant (new technology, changes in political or commercial

**Figure 5.** Knowledge and Ignorance of Humans (Ayyub, 2000),

SME`s are crucial for giving credible data for the risk model. If SMEs are not part of the risk management study, the results will not be trusted, causing this a failure of the process, the PM and the team.
