Integrated project planning and scheduling is a hot research topic that has provided a blueprint of project's success that is based on uninterrupted completion of a project. However, in real production, the project environment changes dynamically because of external and internal fluctuations which create interruptions to the process, due to machine breakdowns, sudden material shortage and so on. These disturbances will mean that the optimal process plan and schedule may become less efficient or even infeasible. Considering preventive maintenance (PM) in the project planning may improve the reliability of the initial plan, but it is conducted at the expense of additional time which may delay the overall process and costs (by the costs of maintenance). Moreover, due to real-world uncertainties, even after efficient PM planning, there are still possibilities of process interruptions in projects. Hence the project managers can be aided by the answers identified in this project to two research questions: (1) when should PM be conducted in the project schedule (frequency of PMs) to reduce the possibilities of interruptions? and (2) in the case of a random interruption, how to revise the existing schedule?

In the era of the fourth generation industry revolution (Industry 4.0), machine learning (ML) techniques enable industries to process data into valuable information to gain proactive production insights for decision-making and process optimization activities. By using these advanced self-learning ML algorithms, project managers can, therefore, make more accurate decisions about PM frequency to enhance the reliability of an initial schedule through the judicious incorporation of PM activities (our first research question). In order to answer the second question, an algorithmic framework will be developed for re-optimizing the existing schedule. The outcome of this project should be helpful for the project managers and maintenance planners tasked with the problem of developing recommended maintenance plans and reactive actions in scheduling real-world complex projects.

Research Aim

To determine the state of the ML techniques in the industry-review data sources from existing literature, industry white papers, etc..

Collect real data (or generate realistic data) about different process interruption and develop ML-based prediction models for predicting interruptions more accurately. Based on those predictions, develop optimization algorithms for scheduling projects while considering PM activities.

Extend the algorithms for re-optimizing the schedules if projects are interrupted by unexpected events.

Effectiveness and validation of the proposed approaches will be driven by verifying findings with selected case studies from different application areas.

Contact:

Contact Dr. Ripon Chakrabortty r.chakrabortty@adfa.edu.au or Dr. Humyun Fuad Rahman md.rahman@unsw.edu.au for further information. Each potential student needs to write a research proposal highlighting research motivation, research problems, research objectives, a brief review of the most relevant literature, proposed methodology and expected outcome. 

School

School of Engineering & IT

Research Area

Technology Decision Making | Optimization & Design | Project Management

Supervisor

PG Project Management, Decision Analytics, and Engineering Science Program Coordinator & Lecturer - Systems Engineering & Decision Analytics Ripon Chakrabortty
PG Project Management, Decision Analytics, and Engineering Science Program Coordinator & Lecturer - Systems Engineering & Decision Analytics