Potential Beneficiaries of the System Summaries and

Abdul Samad Shibghatullah 140 30052008 DA CA1 CA2 CAn reqDriver reqDriver reqDriver respond respond respond detailsSpec detailsSpec detailsSpec beginMatching beginMatching noMatch noMatch findingNearestMatched noMatch findingNearestMatched reserved acceptMatch Figure 6.3: The Sequence of Messages for More than One Round of Rescheduling

6.5 Potential Beneficiaries of the System

The CRSMAS that is informed by this research could be a useful decision making tool for several audiences. Such audiences are Operation Manager, Supervisor, Scheduler, and Researcher. The Operation Manager is a person who is responsible for overseeing the whole bus operation. The Operation Manager can use CRSMAS as a planning tool that is capable of simulating different UE. The simulation results will help the Operation Manager in planning things such as the number of crews needed in everyday operation, and the estimated costs caused by disruption. The Supervisor is a person who is directly involved in day-to-day operations, handling UE, and has to make quick decisions in handling such events. CRSMAS could help the Supervisor in making quick decisions to replace unavailable crew. The Scheduler is a person who is responsible for constructing the predetermined schedules. CRSMAS Abdul Samad Shibghatullah 141 30052008 could help the scheduler to construct robust crew schedules that support the rescheduling process. Section 6.3 discusses some of the limitations in crew schedules that reduce the possibility of successful rescheduling and proposes some actions that could increase successful rescheduling. The Researcher who is searching for tools or techniques to improve bus or public transport services can also use CRSMAS to simulate the rescheduling process using MAS. The simulation results can help researchers to understand the capability of MAS in rescheduling, the effect of UE to schedules, and identify factors in schedules that support rescheduling. The potential beneficiaries mentioned above are those who are involved in bus operation. However, the tool also could be of benefit to other domains of scheduling where humans are involved such as workforce scheduling, nurse rostering, truck driver scheduling and air crew scheduling.

6.6 Summaries and

Conclusions This chapter presents the analysis and evaluation of the results from both the Single Event and Multiple Event experiments. The chapter set two criteria for analysis and evaluation which are rescheduling capability and rescheduling speed. Rescheduling capability was measured using the matching percentage and rescheduling speed was measured by the time taken to perform the rescheduling. For the first criterion, CRSMAS is capable of performing rescheduling but the results are still not satisfactory 44.7 because of weaknesses and limitations in CRSMAS itself, limitations in crew schedules, and the nature of some of the UE which can be lengthy. CRSMAS’s weaknesses and limitations are that it cannot perform more than one round of rescheduling automatically and it cannot reschedule different events at the same time. The limitations in crew schedules are the small number of dutiescrews, duties are not equally spread, and the relief time is not long enough. For the rescheduling speed, the ability of CRSMAS is satisfactory because it can perform rescheduling with the average 1.45 seconds. What is still missing is how we can increase the CRSMAS’s capability for rescheduling. It is hoped that the proposed modifications to CRSMAS, as mentioned in Section 6.4, would increase the matching percentage. Section 6.3 also suggests ways to deal with the limitations in crew schedules and the lengthy period of some UE. From Abdul Samad Shibghatullah 142 30052008 this analysis and evaluation it has been learned that the successful of rescheduling is not only dependent on the tool but also other factors, such as the design of crew schedules. If crew schedules are designed with the criteria suggested in Section 6.3 taken into account, then it will increase the probability of successful rescheduling. One limitation that has been found with MAS is that MAS is not capable of rescheduling for different events at the same time because there are different rules of rescheduling for different events. In short, in this research the research aim was achieved although not 100, which was to reduce the effect of UE to crew schedules. However, there are many potential ways left to maximise the achievement of the aim such as to increase the successful matches, and to design crew schedules that support rescheduling. The chapter suggests potential beneficiaries of the system such as the Operation Manager, Supervisor, Scheduler and Researcher in public transport. The next chapter summarises the whole research and concludes the findings, and suggests further research. Abdul Samad Shibghatullah 143 30052008 Chapter Seven: Summary and Conclusions

7.1 Introduction