Job Shop Scheduling Problem (JSSP)

Problem

In the Job Shop Scheduling Problem (JSSP), a set of jobs has to be processed on every machine in the shop. The jobs consist in ordered sequences of tasks (called activities). An activity represents the processing of the job on one of the machines and has a given processing time. Each job has one activity per machine, and each activity can only start when the previous one has ended. Each machine can only process one activity at a time. The goal is to find a sequence of jobs that minimizes the makespan: the time when all jobs have been processed.

Principles learned

Data

The instances provided follow the Taillard format. The format of the data files is as follows:

  • First line: number of jobs, number of machines, seed used to generate the instance, upper and lower bounds previously found.
  • For each job: the processing time on each machine (given in the processing order).
  • For each job: the processing order (ordered list of visited machines).

Program

The Hexaly model for the Job Shop Scheduling Problem (JSSP) uses interval decision variables to model the time ranges of the activities. We constrain the length of each interval to correspond to the activity’s processing time. We can then write the precedence constraints: for each job, each activity of this job must start after the end of the activity processed by the previous machine.

In addition to the interval decisions representing the time ranges of the activities, we also use list decision variables. As in the Flow Shop example, a list models the ordering of activities on a machine. Using the ‘count’ operator to constrain the size of the lists, we ensure that each machine processes each job. The disjunctive resource constraints — each machine can only process one activity at a time — can be formulated as follows: for all i, the activity processed in position i+1 must start after the end of the activity processed in position i. To model these constraints, we pair up the interval decisions (the time ranges) with the list decisions (the job orderings). We write a lambda function expressing the relationship between two consecutive activities. This function is used within a variadic ‘and’ operator over all activities processed by each machine.

The objective consists in minimizing the makespan, which is the time when all the activities have been processed.

Execution
hexaly jobshop.hxm inFileName=instances/ft10.txt [outFileName=] [hxTimeLimit=]
use io;

/* Read instance data. The input files follow the "Taillard" format */
function input() {
    local usage = "Usage: hexaly jobshop.hxm inFileName=instanceFile "
            + "[outFileName=outputFile] [hxTimeLimit=timeLimit]";
    if (inFileName == nil) throw usage;

    inFile = io.openRead(inFileName);
    inFile.readln();
    nbJobs = inFile.readInt();
    nbMachines = inFile.readInt();
    inFile.readln();
    inFile.readln();
    // Processing times for each job on each machine (given in the processing order)
    processingTimesInProcessingOrder[j in 0...nbJobs][m in 0...nbMachines] = inFile.readInt();
    inFile.readln();
    for [j in 0...nbJobs][k in 0...nbMachines] {
        local m = inFile.readInt() - 1;
        // Processing order of machines for each job
        machineOrder[j][k] = m;
        // Reorder processing times: processingTime[j][m] is the processing time of the
        // task of job j that is processed on machine m
        processingTime[j][m] = processingTimesInProcessingOrder[j][k];
    }
    inFile.close();

    // Trivial upper bound for the start times of the tasks
    maxStart = sum[j in 0...nbJobs][m in 0...nbMachines] (processingTime[j][m]);
}

/* Declare the optimization model */
function model() {
    // Interval decisions: time range of each task
    // tasks[j][m] is the interval of time of the task of job j which is processed
    // on machine m
    tasks[j in 0...nbJobs][m in 0...nbMachines] <- interval(0, maxStart);

    // Task duration constraints
    for [j in 0...nbJobs][m in 0...nbMachines]
        constraint length(tasks[j][m]) == processingTime[j][m];

    // Precedence constraints between the tasks of a job
    for [j in 0...nbJobs][k in 0...nbMachines-1]
        constraint tasks[j][machineOrder[j][k]] < tasks[j][machineOrder[j][k + 1]];

    // Sequence of tasks on each machine
    jobsOrder[m in 0...nbMachines] <- list(nbJobs);

    for [m in 0...nbMachines] {
        // Each job has a task scheduled on each machine
        constraint count(jobsOrder[m]) == nbJobs;

        // Disjunctive resource constraints between the tasks on a machine
        constraint and(0...nbJobs-1,
                i => tasks[jobsOrder[m][i]][m] < tasks[jobsOrder[m][i + 1]][m]);
    }

    // Minimize the makespan: end of the last task of the last job
    makespan <- max[j in 0...nbJobs] (end(tasks[j][machineOrder[j][nbMachines - 1]]));
    minimize makespan;
}

/* Parameterize the solver */
function param() {
    if (hxTimeLimit == nil) hxTimeLimit = 60;
}

/* Write the solution in a file with the following format:
 *  - for each machine, the job sequence */
function output() {
    if (outFileName != nil) {
        outFile = io.openWrite(outFileName);
        println("Solution written in file ", outFileName);
        for [m in 0...nbMachines]
            outFile.println[j in 0...nbJobs](jobsOrder[m].value[j], " ");
    }
}
Execution (Windows)
set PYTHONPATH=%HX_HOME%\bin\python
python jobshop.py instances\ft10.txt
Execution (Linux)
export PYTHONPATH=/opt/hexaly_13_0/bin/python
python jobshop.py instances/ft10.txt
import hexaly.optimizer
import sys


# The input files follow the "Taillard" format
def read_instance(filename):
    with open(filename) as f:
        lines = f.readlines()

    first_line = lines[1].split()
    # Number of jobs
    nb_jobs = int(first_line[0])
    # Number of machines
    nb_machines = int(first_line[1])

    # Processing times for each job on each machine (given in the processing order)
    processing_times_in_processing_order = [[int(lines[i].split()[j])
                                             for j in range(nb_machines)]
                                            for i in range(3, 3 + nb_jobs)]

    # Processing order of machines for each job
    machine_order = [[int(lines[i].split()[j]) - 1 for j in range(nb_machines)]
                     for i in range(4 + nb_jobs, 4 + 2 * nb_jobs)]

    # Reorder processing times: processing_time[j][m] is the processing time of the
    # task of job j that is processed on machine m
    processing_time = [[processing_times_in_processing_order[j][machine_order[j].index(m)]
                        for m in range(nb_machines)]
                       for j in range(nb_jobs)]

    # Trivial upper bound for the start times of the tasks
    max_start = sum(sum(processing_time[j]) for j in range(nb_jobs))

    return nb_jobs, nb_machines, processing_time, machine_order, max_start


def main(instance_file, output_file, time_limit):
    nb_jobs, nb_machines, processing_time, machine_order, max_start = read_instance(instance_file)

    with hexaly.optimizer.HexalyOptimizer() as optimizer:
        #
        # Declare the optimization model
        #
        model = optimizer.model

        # Interval decisions: time range of each task
        # tasks[j][m] is the interval of time of the task of job j which is processed
        # on machine m
        tasks = [[model.interval(0, max_start) for m in range(nb_machines)]
                 for j in range(nb_jobs)]

        # Task duration constraints
        for j in range(nb_jobs):
            for m in range(0, nb_machines):
                model.constraint(model.length(tasks[j][m]) == processing_time[j][m])

        # Create an Hexaly array in order to be able to access it with "at" operators
        task_array = model.array(tasks)

        # Precedence constraints between the tasks of a job
        for j in range(nb_jobs):
            for k in range(nb_machines - 1):
                model.constraint(
                    tasks[j][machine_order[j][k]] < tasks[j][machine_order[j][k + 1]])

        # Sequence of tasks on each machine
        jobs_order = [model.list(nb_jobs) for m in range(nb_machines)]

        for m in range(nb_machines):
            # Each job has a task scheduled on each machine
            sequence = jobs_order[m]
            model.constraint(model.eq(model.count(sequence), nb_jobs))

            # Disjunctive resource constraints between the tasks on a machine
            sequence_lambda = model.lambda_function(
                lambda i: model.lt(model.at(task_array, sequence[i], m),
                                   model.at(task_array, sequence[i + 1], m)))
            model.constraint(model.and_(model.range(0, nb_jobs - 1), sequence_lambda))

        # Minimize the makespan: end of the last task of the last job
        makespan = model.max([model.end(tasks[j][machine_order[j][nb_machines - 1]])
                             for j in range(nb_jobs)])
        model.minimize(makespan)

        model.close()

        # Parameterize the optimizer
        optimizer.param.time_limit = time_limit

        optimizer.solve()

        #
        # Write the solution in a file with the following format:
        # - for each machine, the job sequence
        #
        if output_file != None:
            final_jobs_order = [list(jobs_order[m].value) for m in range(nb_machines)]
            with open(output_file, "w") as f:
                print("Solution written in file ", output_file)
                for m in range(nb_machines):
                    for j in range(nb_jobs):
                        f.write(str(final_jobs_order[m][j]) + " ")
                    f.write("\n")


if __name__ == '__main__':
    if len(sys.argv) < 2:
        print("Usage: python jobshop.py instance_file [output_file] [time_limit]")
        sys.exit(1)

    instance_file = sys.argv[1]
    output_file = sys.argv[2] if len(sys.argv) >= 3 else None
    time_limit = int(sys.argv[3]) if len(sys.argv) >= 4 else 60
    main(instance_file, output_file, time_limit)
Compilation / Execution (Windows)
cl /EHsc jobshop.cpp -I%HX_HOME%\include /link %HX_HOME%\bin\hexaly130.lib
jobshop instances\ft10.txt
Compilation / Execution (Linux)
g++ jobshop.cpp -I/opt/hexaly_13_0/include -lhexaly130 -lpthread -o jobshop
./jobshop instances/ft10.txt
#include "optimizer/hexalyoptimizer.h"
#include <algorithm>
#include <fstream>
#include <iostream>
#include <limits>
#include <numeric>
#include <vector>

using namespace hexaly;
using namespace std;

class Jobshop {
private:
    // Number of jobs
    int nbJobs;
    // Number of machines
    int nbMachines;
    // Processing order of machines for each job
    vector<vector<int>> machineOrder;
    // Processing time on each machine for each job (given in the machine order)
    vector<vector<int>> processingTime;
    // Trivial upper bound for the start times of the tasks
    int maxStart;

    // Hexaly Optimizer
    HexalyOptimizer optimizer;
    // Decision variables: time range of each task
    vector<vector<HxExpression>> tasks;
    // Decision variables: sequence of tasks on each machine
    vector<HxExpression> jobsOrder;
    // Objective = minimize the makespan: end of the last task of the last job
    HxExpression makespan;

public:
    Jobshop() : optimizer() {}

    // The input files follow the "Taillard" format
    void readInstance(const string& fileName) {
        ifstream infile;
        infile.exceptions(ifstream::failbit | ifstream::badbit);
        infile.open(fileName.c_str());

        infile.ignore(numeric_limits<streamsize>::max(), '\n');
        infile >> nbJobs;
        infile >> nbMachines;
        infile.ignore(numeric_limits<streamsize>::max(), '\n');

        // Processing times for each job on each machine (given in the processing order)
        infile.ignore(numeric_limits<streamsize>::max(), '\n');
        vector<vector<int>> processingTimesInProcessingOrder = vector<vector<int>>(nbJobs, vector<int>(nbMachines));
        for (int j = 0; j < nbJobs; ++j) {
            for (int m = 0; m < nbMachines; ++m) {
                infile >> processingTimesInProcessingOrder[j][m];
            }
        }

        // Processing order of machines for each job
        infile.ignore(numeric_limits<streamsize>::max(), '\n');
        infile.ignore(numeric_limits<streamsize>::max(), '\n');
        machineOrder.resize(nbJobs);
        for (int j = 0; j < nbJobs; ++j) {
            machineOrder[j].resize(nbMachines);
            for (int m = 0; m < nbMachines; ++m) {
                int x;
                infile >> x;
                machineOrder[j][m] = x - 1;
            }
        }

        // Reorder processing times: processingTime[j][m] is the processing time of the
        // task of job j that is processed on machine m
        for (int j = 0; j < nbJobs; ++j) {
            processingTime.resize(nbJobs);
            for (int m = 0; m < nbMachines; ++m) {
                processingTime[j].resize(nbMachines);
                vector<int>::iterator findM = find(machineOrder[j].begin(), machineOrder[j].end(), m);
                unsigned int k = distance(machineOrder[j].begin(), findM);
                processingTime[j][m] = processingTimesInProcessingOrder[j][k];
            }
        }

        // Trivial upper bound for the start times of the tasks
        maxStart = 0;
        for (int j = 0; j < nbJobs; ++j)
            maxStart += accumulate(processingTime[j].begin(), processingTime[j].end(), 0);

        infile.close();
    }

    void solve(int timeLimit) {
        // Declare the optimization model
        HxModel model = optimizer.getModel();

        // Interval decisions: time range of each task
        // tasks[j][m] is the interval of time of the task of job j which is processed on machine m
        tasks.resize(nbJobs);
        for (unsigned int j = 0; j < nbJobs; ++j) {
            tasks[j].resize(nbMachines);
            for (unsigned int m = 0; m < nbMachines; ++m) {
                tasks[j][m] = model.intervalVar(0, maxStart);

                // Task duration constraints
                model.constraint(model.length(tasks[j][m]) == processingTime[j][m]);
            }
        }

        // Create an Hexaly array in order to be able to access it with "at" operators
        HxExpression taskArray = model.array();
        for (int j = 0; j < nbJobs; ++j) {
            taskArray.addOperand(model.array(tasks[j].begin(), tasks[j].end()));
        }

        // Precedence constraints between the tasks of a job
        for (int j = 0; j < nbJobs; ++j) {
            for (int k = 0; k < nbMachines - 1; ++k) {
                model.constraint(tasks[j][machineOrder[j][k]] < tasks[j][machineOrder[j][k + 1]]);
            }
        }

        // Sequence of tasks on each machine
        jobsOrder.resize(nbMachines);
        for (int m = 0; m < nbMachines; ++m) {
            jobsOrder[m] = model.listVar(nbJobs);
        }

        for (int m = 0; m < nbMachines; ++m) {
            // Each job has a task scheduled on each machine
            HxExpression sequence = jobsOrder[m];
            model.constraint(model.eq(model.count(sequence), nbJobs));

            // Disjunctive resource constraints between the tasks on a machine
            HxExpression sequenceLambda = model.createLambdaFunction([&](HxExpression i) {
                return model.at(taskArray, sequence[i], m) < model.at(taskArray, sequence[i + 1], m);
            });
            model.constraint(model.and_(model.range(0, nbJobs - 1), sequenceLambda));
        }

        // Minimize the makespan: end of the last task of the last job
        makespan = model.max();
        for (int j = 0; j < nbJobs; ++j) {
            makespan.addOperand(model.end(tasks[j][machineOrder[j][nbMachines - 1]]));
        }
        model.minimize(makespan);

        model.close();

        // Parameterize the optimizer
        optimizer.getParam().setTimeLimit(timeLimit);

        optimizer.solve();
    }

    /* Write the solution in a file with the following format:
     *  - for each machine, the job sequence */
    void writeSolution(const string& fileName) {
        ofstream outfile;
        outfile.exceptions(ofstream::failbit | ofstream::badbit);
        outfile.open(fileName.c_str());
        cout << "Solution written in file " << fileName << endl;

        for (int m = 0; m < nbMachines; ++m) {
            HxCollection finalJobsOrder = jobsOrder[m].getCollectionValue();
            for (int j = 0; j < nbJobs; ++j) {
                outfile << finalJobsOrder.get(j) << " ";
            }
            outfile << endl;
        }
        outfile.close();
    }
};

int main(int argc, char** argv) {
    if (argc < 2) {
        cout << "Usage: jobshop instanceFile [outputFile] [timeLimit]" << endl;
        exit(1);
    }

    const char* instanceFile = argv[1];
    const char* outputFile = argc > 2 ? argv[2] : NULL;
    const char* strTimeLimit = argc > 3 ? argv[3] : "60";

    Jobshop model;
    try {
        model.readInstance(instanceFile);
        const int timeLimit = atoi(strTimeLimit);
        model.solve(timeLimit);
        if (outputFile != NULL)
            model.writeSolution(outputFile);
        return 0;
    } catch (const exception& e) {
        cerr << "An error occurred: " << e.what() << endl;
        return 1;
    }
}
Compilation / Execution (Windows)
copy %HX_HOME%\bin\Hexaly.NET.dll .
csc Jobshop.cs /reference:Hexaly.NET.dll
Jobshop instances\ft10.txt
using System;
using System.IO;
using Hexaly.Optimizer;

public class Jobshop : IDisposable
{
    // Number of jobs
    private int nbJobs;

    // Number of machines
    private int nbMachines;

    // Processing order of machines for each job
    private int[,] machineOrder;

    // Processing time on each machine for each job (given in the machine order)
    private long[,] processingTime;

    // Trivial upper bound for the start times of the tasks
    private long maxStart;

    // Hexaly Optimizer
    private HexalyOptimizer optimizer;

    // Decision variables: time range of each task
    private HxExpression[,] tasks;

    // Decision variables: sequence of tasks on each machine
    private HxExpression[] jobsOrder;

    // Objective = minimize the makespan: end of the last task of the last job
    private HxExpression makespan;

    public Jobshop()
    {
        optimizer = new HexalyOptimizer();
    }

    // The input files follow the "Taillard" format
    public void ReadInstance(string fileName)
    {
        using (StreamReader input = new StreamReader(fileName))
        {
            input.ReadLine();
            string[] splitted = input.ReadLine().Split(' ');
            nbJobs = int.Parse(splitted[0]);
            nbMachines = int.Parse(splitted[1]);

            // Processing times for each job on each machine (given in the processing order)
            input.ReadLine();
            long[,] processingTimesInProcessingOrder = new long[nbJobs, nbMachines];
            for (int j = 0; j < nbJobs; ++j)
            {
                splitted = input.ReadLine().Trim().Split(' ');
                for (int m = 0; m < nbMachines; ++m)
                    processingTimesInProcessingOrder[j, m] = long.Parse(splitted[m]);
            }

            // Processing order of machines for each job
            input.ReadLine();
            machineOrder = new int[nbJobs, nbMachines];
            for (int j = 0; j < nbJobs; ++j)
            {
                splitted = input.ReadLine().Trim().Split(' ');
                for (int m = 0; m < nbMachines; ++m)
                    machineOrder[j, m] = int.Parse(splitted[m]) - 1;
            }

            // Reorder processing times: processingTime[j, m] is the processing time of the
            // task of job j that is processed on machine m
            processingTime = new long[nbJobs, nbMachines];
            // Trivial upper bound for the start times of the tasks
            maxStart = 0;
            for (int j = 0; j < nbJobs; ++j)
            {
                for (int m = 0; m < nbMachines; ++m)
                {
                    int machineIndex = nbMachines;
                    for (int k = 0; k < nbMachines; ++k)
                    {
                        if (machineOrder[j, k] == m)
                        {
                            machineIndex = k;
                            break;
                        }
                    }
                    processingTime[j, m] = processingTimesInProcessingOrder[j, machineIndex];
                    maxStart += processingTime[j, m];
                }
            }
        }
    }

    public void Dispose()
    {
        optimizer.Dispose();
    }

    public void Solve(int timeLimit)
    {
        // Declare the optimization model
        HxModel model = optimizer.GetModel();

        // Interval decisions: time range of each task
        // tasks[j][m] is the interval of time of the task of job j which is processed on machine m
        tasks = new HxExpression[nbJobs, nbMachines];
        for (int j = 0; j < nbJobs; ++j)
        {
            for (int m = 0; m < nbMachines; ++m)
            {
                tasks[j, m] = model.Interval(0, maxStart);

                // Task duration constraints
                model.Constraint(model.Length(tasks[j, m]) == processingTime[j, m]);
            }
        }

        // Create a HexalyOptimizer array in order to be able to access it with "at" operators
        HxExpression taskArray = model.Array(tasks);

        // Precedence constraints between the tasks of a job
        for (int j = 0; j < nbJobs; ++j)
        {
            for (int k = 0; k < nbMachines - 1; ++k)
            {
                model.Constraint(tasks[j, machineOrder[j, k]] < tasks[j, machineOrder[j, k + 1]]);
            }
        }

        // Sequence of tasks on each machine
        jobsOrder = new HxExpression[nbMachines];
        for (int m = 0; m < nbMachines; ++m)
            jobsOrder[m] = model.List(nbJobs);

        for (int m = 0; m < nbMachines; ++m)
        {
            // Each job has a task scheduled on each machine
            HxExpression sequence = jobsOrder[m];
            model.Constraint(model.Count(sequence) == nbJobs);

            // Disjunctive resource constraints between the tasks on a machine
            HxExpression sequenceLambda = model.LambdaFunction(
                i => taskArray[sequence[i], m] < taskArray[sequence[i + 1], m]
            );
            model.Constraint(model.And(model.Range(0, nbJobs - 1), sequenceLambda));
        }

        // Minimize the makespan: end of the last task of the last job
        makespan = model.Max();
        for (int j = 0; j < nbJobs; ++j)
            makespan.AddOperand(model.End(tasks[j, machineOrder[j, nbMachines - 1]]));
        model.Minimize(makespan);

        model.Close();

        // Parameterize the optimizer
        optimizer.GetParam().SetTimeLimit(timeLimit);

        optimizer.Solve();
    }

    /* Write the solution in a file with the following format:
     *  - for each machine, the job sequence */
    public void WriteSolution(string fileName)
    {
        using (StreamWriter output = new StreamWriter(fileName))
        {
            Console.WriteLine("Solution written in file " + fileName);
            for (int m = 0; m < nbMachines; ++m)
            {
                HxCollection finalJobsOrder = jobsOrder[m].GetCollectionValue();
                for (int i = 0; i < nbJobs; ++i)
                {
                    int j = (int)finalJobsOrder.Get(i);
                    output.Write(j + " ");
                }
                output.WriteLine();
            }
        }
    }

    public static void Main(string[] args)
    {
        if (args.Length < 1)
        {
            Console.WriteLine("Usage: Jobshop instanceFile [outputFile] [timeLimit]");
            System.Environment.Exit(1);
        }

        string instanceFile = args[0];
        string outputFile = args.Length > 1 ? args[1] : null;
        string strTimeLimit = args.Length > 2 ? args[2] : "60";

        using (Jobshop model = new Jobshop())
        {
            model.ReadInstance(instanceFile);
            model.Solve(int.Parse(strTimeLimit));
            if (outputFile != null)
                model.WriteSolution(outputFile);
        }
    }
}
Compilation / Execution (Windows)
javac Jobshop.java -cp %HX_HOME%\bin\hexaly.jar
java -cp %HX_HOME%\bin\hexaly.jar;. Jobshop instances\ft10.txt
Compilation / Execution (Linux)
javac Jobshop.java -cp /opt/hexaly_13_0/bin/hexaly.jar
java -cp /opt/hexaly_13_0/bin/hexaly.jar:. Jobshop instances/ft10.txt
import java.util.*;
import java.io.*;
import com.hexaly.optimizer.*;

public class Jobshop {
    // Number of jobs
    private int nbJobs;
    // Number of machines
    private int nbMachines;
    // Processing time on each machine for each job (given in the machine order)
    private long[][] processingTime;
    // Processing order of machines for each job
    private int[][] machineOrder;
    // Trivial upper bound for the start times of the tasks
    private long maxStart;

    // Hexaly Optimizer
    final HexalyOptimizer optimizer;
    // Decision variables: time range of each task
    private HxExpression[][] tasks;
    // Decision variables: sequence of tasks on each machine
    private HxExpression[] jobsOrder;
    // Objective = minimize the makespan: end of the last task of the last job
    private HxExpression makespan;

    public Jobshop(HexalyOptimizer optimizer) throws IOException {
        this.optimizer = optimizer;
    }

    // The input files follow the "Taillard" format
    public void readInstance(String fileName) throws IOException {
        try (Scanner input = new Scanner(new File(fileName))) {
            input.nextLine();
            nbJobs = input.nextInt();
            nbMachines = input.nextInt();

            input.nextLine();
            input.nextLine();
            // Processing times for each job on each machine (given in the processing order)
            long[][] processingTimesInProcessingOrder = new long[nbJobs][nbMachines];
            for (int j = 0; j < nbJobs; ++j) {
                for (int m = 0; m < nbMachines; ++m) {
                    processingTimesInProcessingOrder[j][m] = input.nextInt();
                }
            }
            // Processing order of machines for each job
            input.nextLine();
            input.nextLine();
            machineOrder = new int[nbJobs][nbMachines];
            for (int j = 0; j < nbJobs; ++j) {
                for (int m = 0; m < nbMachines; ++m) {
                    machineOrder[j][m] = input.nextInt() - 1;
                }
            }
            // Reorder processing times: processingTime[j][m] is the processing time of the
            // task of job j that is processed on machine m
            processingTime = new long[nbJobs][nbMachines];
            // Trivial upper bound for the start times of the tasks
            maxStart = 0;
            for (int j = 0; j < nbJobs; ++j) {
                for (int m = 0; m < nbMachines; ++m) {
                    int machineIndex = nbMachines;
                    for (int k = 0; k < nbMachines; ++k) {
                        if (machineOrder[j][k] == m) {
                            machineIndex = k;
                            break;
                        }
                    }
                    processingTime[j][m] = processingTimesInProcessingOrder[j][machineIndex];
                    maxStart += processingTime[j][m];
                }
            }
        }
    }

    public void solve(int timeLimit) {
        // Declare the optimization model
        HxModel model = optimizer.getModel();

        // Interval decisions: time range of each task
        // tasks[j][m] is the interval of time of the task of job j which is processed
        // on machine m
        tasks = new HxExpression[nbJobs][nbMachines];
        for (int j = 0; j < nbJobs; ++j) {
            for (int m = 0; m < nbMachines; ++m) {
                tasks[j][m] = model.intervalVar(0, maxStart);

                // Task duration constraints
                model.constraint(model.eq(model.length(tasks[j][m]), processingTime[j][m]));
            }
        }

        // Create a HexalyOptimizer array in order to be able to access it with "at"
        // operators
        HxExpression taskArray = model.array(tasks);

        // Precedence constraints between the tasks of a job
        for (int j = 0; j < nbJobs; ++j) {
            for (int k = 0; k < nbMachines - 1; ++k) {
                model.constraint(model.lt(tasks[j][machineOrder[j][k]], tasks[j][machineOrder[j][k + 1]]));
            }
        }

        // Sequence of tasks on each machine
        jobsOrder = new HxExpression[nbMachines];
        for (int m = 0; m < nbMachines; ++m) {
            jobsOrder[m] = model.listVar(nbJobs);
        }

        for (int m = 0; m < nbMachines; ++m) {
            // Each job has a task scheduled on each machine
            HxExpression sequence = jobsOrder[m];
            model.constraint(model.eq(model.count(sequence), nbJobs));

            // Disjunctive resource constraints between the tasks on a machine
            HxExpression mExpr = model.createConstant(m);
            HxExpression sequenceLambda = model
                    .lambdaFunction(i -> model.lt(model.at(taskArray, model.at(sequence, i), mExpr),
                            model.at(taskArray, model.at(sequence, model.sum(i, 1)), mExpr)));
            model.constraint(model.and(model.range(0, nbJobs - 1), sequenceLambda));
        }

        // Minimize the makespan: end of the last task of the last job
        makespan = model.max();
        for (int j = 0; j < nbJobs; ++j) {
            makespan.addOperand(model.end(tasks[j][machineOrder[j][nbMachines - 1]]));
        }
        model.minimize(makespan);

        model.close();

        // Parameterize the optimizer
        optimizer.getParam().setTimeLimit(timeLimit);

        optimizer.solve();
    }

    /*
     * Write the solution in a file with the following format:
     * - for each machine, the job sequence
     */
    public void writeSolution(String fileName) throws IOException {
        try (PrintWriter output = new PrintWriter(fileName)) {
            System.out.println("Solution written in file " + fileName);

            for (int m = 0; m < nbMachines; ++m) {
                HxCollection finalJobsOrder = jobsOrder[m].getCollectionValue();
                for (int i = 0; i < nbJobs; ++i) {
                    int j = Math.toIntExact(finalJobsOrder.get(i));
                    output.write(j + " ");
                }
                output.write("\n");
            }
        }
    }

    public static void main(String[] args) {
        if (args.length < 1) {
            System.out.println("Usage: java Jobshop instanceFile [outputFile] [timeLimit]");
            System.exit(1);
        }

        String instanceFile = args[0];
        String outputFile = args.length > 1 ? args[1] : null;
        String strTimeLimit = args.length > 2 ? args[2] : "60";

        try (HexalyOptimizer optimizer = new HexalyOptimizer()) {
            Jobshop model = new Jobshop(optimizer);
            model.readInstance(instanceFile);
            model.solve(Integer.parseInt(strTimeLimit));
            if (outputFile != null) {
                model.writeSolution(outputFile);
            }
        } catch (Exception ex) {
            System.err.println(ex);
            ex.printStackTrace();
            System.exit(1);
        }
    }
}

Results

Hexaly Optimizer improves the literature’s best known solution by an average of 7.4% within 1 minute of running time on large-scale instances of the Job Shop Scheduling Problem (JSSP) [1]. Our Job Shop Scheduling (JSSP) benchmark page illustrates how Hexaly Optimizer outperforms traditional general-purpose optimization solvers, like IBM ILOG CP Optimizer, Google OR-Tools, Gurobi, and IBM ILOG Cplex, on this challenging problem.


[1] Giacomo Da Col, Erich C. Teppan, Industrial-size job shop scheduling with constraint programming, Operations Research Perspectives, Volume 9, 2022.