Context and challenges
This client, a major player in the nuclear industry, manages various projects, which require digitalized project management tools. Specifically in the context of project planning of large-scale projects, this kind of organisations require efficient planning solutions that can meet various project objectives.
These objectives include to plan the activities, while minimizing the project delays and maximizing the usage of available resources. It can also include timely completion of the project and its milestones or optimization of constraints. For instance, the client required optimization of its planning for planned maintenance shutdowns, where the number of cycles for switching of water levels need to be minimized.
Delays can have serious financial consequences for this client, and it required indeed optimisation of its activities under constraints. Moreover, each project is unique in terms of practical constraints to integrate (e.g., minimization of frequency of changing of water levels), while also taking into account expert inputs (e.g. specification of task precedence - prerequisites - or combination of resources required).
In the above context, this client appointed Assystem to provide a solution to optimise project schedules.
Assystem provided a customized constraint-based engine solution, that can be tailored to comply with any kind of project rules as per the client’s need.
In-house software development of a fast and efficient python package, including:
- Architecture specification, including schemes of the hierarchy of classes and modules, etc.
- Conception of a dynamic scoring function to compute the dynamic priority of tasks with a local and global view of the optimization
- KPIs including milestones summary, task delays and conflicts, scalable Gantt chart, etc.
- Testing models to benchmark/challenge the solution
- Unit/integration tests including white/black box testing and system testing
The solution includes the following features:
- Flexibility to tailor as per project and the client’s specific rules and constraints
- Optimization based on a rule-based engine to include all possible project constraints
- Ability to override specific system decisions by adding user interaction, enabling to customize to solution to the client needs
- Inclusion of artificial intelligence algorithms (genetic algorithms) and real life uncertainties with probabilistic models
- Inclusion of data visualization (graphics, dashboards) to highlight conflicts in initial planning and the way to resolve it
- Correction of existing schedules with the integration of standard data formats from popular software packages such as Planisware, MS Project or CosmoTech
- Optimization of project plans by providing information without incoherencies and conflicts regarding the initial planning
- Decision support using a user interface with relevant KPIs
- Ability to include project constraints without increasing the mathematical complexity
- Reduction of project delays and maximise resource usage (equipment, human resources) thanks to optimized planning, useful indicators (KPI of the project) and forecasting project plans
- Computational efficiency to provide solutions for large-scale industrial projects