An Expert system or known as knowledge based system is primarily an artificial intelligence technique that uses expert knowledge to solve complex decision problems and can be viewed as a computer simulation of a human knowledge (Ford, 1991). An expert system differs from human expertise which allows integration of knowledge from different sources and makes it available to a large number of less skilled people skilled (Finlay and Dix, 1996 and Ford, 1991).
“Such a system may completely fulfil a function that normally requires human expertise, or it may play the role of an assistance to a human decision makers” (Jackson, 1990, p3). “Expert systems not only possess human knowledge in the form of coded tables, databases, and programmed logic, but are coming closer and closer to adequately and truly representing human systems that think. Built to include the ever popular modular structure, expert systems can be refined and improved, just as a human’s thoughts can”, (Ruchelsman, 2004, p. 2)
Moreover, expert systems are usually used commercially because they are less expensive in terms of searching for knowledge and also in training new unskilled people (Finlay and Dix, 1996 and Ford, 1991). Finally, expert systems equip the users with unbiased response and thus provide correct solution to problems. In order to be certain that the problem will be solved appropriately by an expert system, two tests should be passed (Finlay and Dix, 1996). Firstly, the problem should fall in one of these categories Secondly, whether the problem can be sufficiently solved using conventional and cheaper techniques?
For example, whether it can be solved statistically? If the answer to both of these questions is no, then we should consider whether the problem justifies the expenses and efforts required to build such systems because building an expert system aims to save costs in the long term. When both tests are successfully passed several phases are followed to build such a system. The first phase is knowledge acquisition which is the most crucial stage in building the system (Finlay and Dix, 1996). Knowledge base acquisition contains both factual and heuristic knowledge (Engelmore & Feigenbaum, 1993).
Factual knowledge is that knowledge which is widely shared, typically called secondary source knowledge found in books or journals While Heuristic knowledge is known as main expert source knowledge rarely discussed, more experiential, more judgmental knowledge of performance than the factual knowledge (Engelmore & Feigenbaum, 1993). This is then followed by the design and development phase. This phase is a highly iterative set of processes in which the designer builds a part of the system and then tests the result (Lukasheh et al. 2001). System knowledge is modified based on test results and modifications takes place after every single test.
The final phase is the evaluation process in which the system is implemented. On one hand it is said that, “Real time traffic information for all major roadways in an urban area is one of the most important pieces of information necessary to produce a dynamic route guidance”(Nual et al. 2002, p. 1). On the other hand, using such data can also be very expensive and not available to all communities (Dahlgren, 2002; Nual et al. 2002). The UK Parliamentary Office of Science and Technology (2002) addressed two main issues of using real time data. This first one was the financing and management issues.
This was because different applications rely on using road side traffic mentoring sensors, communication equipments, digital UK maps and other technologies, thus it is difficult to quantify the infrastructure cost for each application individually. Moreover, extra costs are needed to upgrade these systems frequently because the technologies which are associated with these systems develop at a rapid pace. As a result of this issue the “Decision-makers may be reluctant to deploy ITS systems without quantitative data on costs and benefits” (UK Parliamentary Office of Science and Technology, 2002), pp4).
The second issue is the process and the efficient use of large amount of real time data which is another challenge. For instance, there is no benefit of the system, if the transmitted information from automated emergency call system to the local emergency service is not transmitted rapidly. In addition to these issues, collecting accurate real time data for traffic conditions is not accurate, and requires much resources and staff (Thompson, 2003). Thompson mentioned that there are currently two techniques used to collect real time data of the congestion on roads: spot speed measurement techniques and vehicle tracing techniques.
The first technique collects the accelerations of vehicles from specific times and points, which according to him, affects the accuracy of estimating the acceleration of the cars. The second technique is not accurate either because it relies on using moving vehicles that are mounted with speed detectors to collect speed data by tracking sample cars randomly. He added that collecting real time data for congested roads also requires special trained survey technicians and expensive equipments. It is also labour intensives and requires trained technicians, in which case extra money needs to be provided (Thompson, 2003).