PLAIO Copilot

How are pharma companies coping in this volatile environment?

Conflicting objectives can cause issues, such as the need to maximize resource usage to keep production costs down, while simultaneously striving for a high customer service rate.

Multi-objective decision-making is complex and often difficult for the human mind to deal with. Without a systematic approach, humans tend to base their decisions on only one of the multiple objectives. Herbert A. Simon, the American political scientist who received the Nobel Prize for economics science in 1978, described this in his decision-making theory that explains bounded rationality and how people make decisions within certain limitations.

When humans leave out important objectives, they are likely to accidentally exclude the optimal solutions that the industry urgently needs in the current environment.

So, what are companies to do?

The focus appears to be simple to set: make more for less and optimize at all costs.

Successful planning and scheduling tend to be complex and include many different aspects. Numerous factors, including market knowledge and personal experience, affect the planner’s decisions. As the schedule grows larger, these decisions become more error-prone and difficult to comprehend. Guidelines may not even lead to a near-optimal solution.

In addition to the more complex situations, the planner also must assign straightforward tasks to the plan, which can be both tedious and time-consuming.

The industry is actively looking at solutions for solving this scheduling problem, without disrupting the manufacturing process. Today, most pharma companies’ schedules are still created and maintained manually by the planner, using less-than-ideal software solutions, often spreadsheet-based.

Problems are usually the predictable result.

Complicated spreadsheets are notoriously difficult to debug and maintain, so users tend to tread very carefully in order not to break something that is working well enough. Version control and lack of history often result in companies relying on a single master plan, without any obvious or easy ways to perform what-if-analysis, or algorithms to suggest improvements to the plans.

Hence the manufacturing plans are usually sub-optimal, difficult to maintain, and hinder necessary responses to changes in the environment.

From manual spreadsheets to automation

Modifying the traditional manufacturing workflow means replacing manual spreadsheet calculations with automation and replacing email communications with a built-in communication tool and functional BI-user interface.

Enhanced production scheduling tools can improve decision-making, save time, and have a significant impact on the bottom line.

This is where the PLAIO Copilot comes in.

For the planner, full control of the schedule is important. PLAIO’s Copilot is therefore designed not to take over, but to suggest better solutions and automate straightforward tasks.

Planners are often faced with choices where there is no obviously correct answer. These are not situations where computers should make autonomous decisions. Instead, the expert planners should always have the final say. A key strength of the Copilot is the ability to give the planner options and suggestions and help clear up obvious and repetitive tasks. The Copilot gives the user the ability to review and pick-and-choose from its suggestions using a simple user interface.

The Copilot considers all important objectives simultaneously and can be used by the planner to overcome the limitations of human decision-making, including bounded rationality.

The Copilot will assist the planner with various data-driven quantitative methods supporting decision-making, ranging from very simple rule-based methods to more advanced artificial intelligence and optimisation-based methods for planning and scheduling. The quantitative methods result in suggestions that are classified in three categories;

• Instant suggestions: The Copilot instantly suggests improvements to the current plan, by performing repetitive and simple time-consuming tasks for the planner using heuristic-based algorithms.
• Coffee-break suggestions: The Copilot suggests improvements to the plan, that take a few minutes to run using heuristic algorithms, search methods, and simple optimization.
• Overnight/weekend: The Copilot thoroughly optimizes the schedule, taking into account all parameters. The result is a brand-new plan that can be compared to the active master plan, in order to evaluate its quality and see if larger improvements can be made. The detailed optimization is also useful for creating long-term plans and running extensive what-if scenarios to facilitate higher-level decision-making.

These methods used by the Copilot will reduce manual labour significantly, saving time and resources, while eliminating various tedious and repetitive tasks. The Copilot will also help improve the planner’s decisions, with more insight into the advantages of different solutions.

In the current business environment, the Copilot can help promote the planner from passenger to pilot.