How to create a model that assigns a time, employee, and equipment to a job?

Context: This is my first time working with Timefold/Optaplanner, so I might have some incorrect understanding of everything.

I'm working on developing a Quarkus Timefold app using/modifying the food packaging example. The goal is to update the model to be able to assign a time, which would be based on the previous job, an employee with the required skill, and required equipment. My model currently has a Job and Line from the example, and I'm trying to add a new Employee class. However, I'm getting this error:

java.lang.IllegalArgumentException: The config (QueuedValuePlacerConfig(ValueSelectorConfig(jobs), ListChangeMoveSelectorConfig(ValueSelectorConfig(null), null))) has no entityClass configured and because there are multiple in the entityClassSet ([class org.acme.foodpackaging.domain.Line, class org.acme.foodpackaging.domain.Job]), it cannot be deduced automatically.

In the job class, I added

@PlanningVariable(valueRangeProviderRefs = "employees")
private Employee employee;

and in the PackagingSchedule class, I added

@ProblemFactCollectionProperty
@ValueRangeProvider(id = "employees")
private List<Employee> employees;

This is my employee class:

package org.acme.foodpackaging.domain;

import java.time.LocalDateTime;
import java.util.*;

public class Employee {

    private Long id;
    private String name;
    private String operator;
    private LocalDateTime startDateTime;
    Set<String> skills;

    // No-arg constructor required for OptaPlanner and Jackson
    public Employee() {
    }

    public Employee(Long id, String name, String operator, LocalDateTime startDateTime) {
        this.id = id;
        this.name = name;
        this.operator = operator;
        this.startDateTime = startDateTime;
        this.skills = null;
    }

    public Employee(Long id, String name, String operator, LocalDateTime startDateTime, Set<String> skills) {
        this.id = id;
        this.name = name;
        this.operator = operator;
        this.startDateTime = startDateTime;
        this.skills = skills;
    }

    @Override
    public String toString() {
        return name;
    }

    // ************************************************************************
    // Getters and setters
    // ************************************************************************

    public Long getId() {
        return id;
    }

    public String getName() {
        return name;
    }

    public String getOperator() {
        return operator;
    }

    public LocalDateTime getStartDateTime() {
        return startDateTime;
    }

    public Set<String> getSkills() { return skills; }

}

I believe this error is because the Job class, which was originally a PlanningEntity with no genuine planning variables, now contains a genuine planning variable. How do I correctly create this model?

Solution

That error message should be improved, but the issue is that currently @PlanningListVariable cannot exist with another @PlanningVariable or @PlanningListVariable. The variable defined by Line is a @PlanningListVariable, which means no other genuine variables can be defined (as noted in the docs: https://timefold.ai/docs/timefold-solver/latest/using-timefold-solver/modeling-planning-problems#planningListVariable

Planning list variable does not yet support all the advanced planning features that work with the chained planning variable. Use a chained planning variable instead of a planning list variable, if you need any of the following planning techniques:

overconstrained planning,

planning entity difficulty comparison or planning value strength comparison,

exhaustive search,

partitioned search,

coexistence with another list or basic planning variable.

Converting the model to chained is not trivial; essentially, Line becomes a problem fact that acts as an anchor, and Job becomes a planning entity that points either to a Job or a Line. I recommend you look at the ordering picking quickstart (https://github.com/TimefoldAI/timefold-quickstarts/tree/stable/use-cases/order-picking) for an example of creating a chained model.