CASE STUDY- I

TASK SCHEDULING

• Presented By : Muhammad Fahad

Sumaya Abdul Rahman
• Manal AlZaidan
• Abdus Samad

• Submitted to : Dr. Adel Elomri

Problem Statement

7 employees

25 tasks
Question 1

• Task Scheduling
• A manager of a company is aiming to scheduling a set of tasks over
his employees. Each task is characterized by a duration (in minutes) and
can be handled by any of the employees (skills independent). Any
task should be allocated to only one employee, it cannot be divided
among many employees.

• Question- Propose a Linear Program that helps providing a “task

number-based Fair” schedule. Task number-based fairness is measured in
the sense of total number of tasks allocated to each employee.
Goal Programming based approach for task number
based fair scheduling of Employees

Decision Variables:
• X ij : is a binary decision variable indicating whether task I is assigned to employee j.
• d1j : is a decision variable representing the overhead deviation for employee j.
• d2j : is a decision variable representing the under-head deviation for employee j.
Objective Function:
• Minimize the sum of overhead and under-head deviation in assigned tasks for each employee
from the goal.

Minimize
Our Approach and Solution
Constraints:
• Each task must be assigned to exactly one employee no task to be split among
employees.

• Each employee can handle a minimum of one task to avoids scenarios where some
employees might be left without any tasks, contributing to work distribution fairness.

• Define deviation variables for each employee's goal:

• Define non-negative deviation variables for overhead:

• Define non-negative deviation variables for under head:

 OPL Code

//Parameters
int m 25 // Number of tasks
int n 7 // Number of employees
range M 1 m
range N 1 n
float goal 3.57 //Average number of Tasks (Number of Tasks/Number of Employees)
//Decision Variables
dvar boolean x // X=1 if task M has been assigned to employee N, otherwise X=0
dvar float d N // overhead Deviation variable (Upper side deviation in assigned tasks from the
goal)
dvar float S N //underhead Deviation variables (Lower side deviation in assigned tasks from the
goal)
 OPL Code

//Objective Function

//Minimize the sum of Upper side and Lower side deviation in assigned tasks for each employee from the goal

minimize sum(j in N) (d[j]+S[j]);

//Constraints
subject to {
// Constraint that each task must be assigned to exactly one employee (cannot split among employees)
C1 : forall (i in M)

sum(j in N) x[j][i] == 1;
// Constraint that each employee will handle a minimum of one task (No one with zero task assignment)

C2 : forall (j in N)

sum(i in M) x[j][i] >= 1;

 OPL Code

// Constraint on deviation variables for each employee's goal

C3 : forall (j in N)
sum(i in M) x[j][i]-d[j]+S[j]==goal;

// Constraint on Non-negative overhead deviation variable

C4 : forall (j in N)
d[j] >= 0;

// Constraint on Non-negative underhead deviation variable

C5 : forall (j in N)

S[j] >= 0;
}
OPL Results
 Graphical Illustration

• - Given that there are 25 tasks and 7 employees, the linear programming solution
optimizes task allocation.
• - The solution assigns 4 tasks to 4 employees and 3 tasks to the remaining 3
employees, ensuring a fair distribution.
• -The result of this allocation sums up to the total of 25 tasks, maintaining task
completeness and fairness.
Total minimized deviation= 3.5
Employees with 3 tasks : D+=[0.5,0.5,0.5,0,0,0,0]
Employees with 4 tasks : D-=[0,0,0,0.5,0.5,0.5,0.5]
GRAPH ILLUSTRATION
0.5 0.5 0.5 0.5
goal=3.5

0.5 0.5 0.5

 Task Scheduling

Question
A manager of a company is aiming to scheduling a set
of tasks over his employees. Each task is characterized by
a duration (in minutes) and can be handled by any of the

2
employees (skills independent). Any task should be
allocated to only one employee, it cannot be divided
among many employees.

Question: Propose a Linear Program that helps providing

a “duration-based Fair” schedule. Duration-based fairness
is measured in the sense of total work duration (in
minutes) allocated to each employee.
Our Approach and Solution
Sets:
• M is the set of tasks, M = {1,2,….,m}
• N is the set of employees, N = {1,2,….,n}
• Decision Variable:
• Xij : is a binary decision variable indicating whether task I is assigned to employee j.
• d1j : is a continuous decision variable representing the overhead deviation for employee j.
• d2j : is a continuous decision variable representing the under head deviation for employee j.
Data:
• dr[M] is an array representing the duration of tasks M.
• Goal is the average duration of tasks assigned to each employee.
• Objective Function:
• Minimize the total deviation from the goal for all employees:

Minimize
Our Approach and Solution
Constraints:
• Each task must be assigned to exactly one employee:

• Each employee must handle a minimum of one task:

• Define deviation variables for each employee's goal:

• Define non-negative deviation variables for over-head:

• Define non-negative deviation variables for under head:

 Graph Illustration
• -There are a total of 25 tasks. These tasks are assigned among 7 employees.

• - Each task has a specific duration, and the durations for the 25 tasks are as follows:
• duration[25] = [17,62,91,21,31,97,57,55,40,52,60,52,73,46,63,46,28,72,10,36,28,74,100,65,11]

• - The average duration chosen of all tasks is 184.

• - According to our Linear Programming (LP) solution:

• - 6 employees have been assigned 183 tasks each and 1 employee has been assigned 183 tasks
• - This results in a deviation of 1 .
 OPL Code

/int n = 7;// Total number of employees

int m = 25;// Total number of tasks

range M = 1..m; // Set of tasks
range N = 1..n; // Set of employees

dvar boolean x[N][M]; // X=1 if task M has been assigned to emloyee N, otherwise X=0
dvar float d1[N]; // overhead Deviation variable (Upper side deviation in assigned tasks from the goal)

dvar float d2[N]; //underhead Deviation variables (Lower side deviation in assigned tasks from the goal)

float dr[M] = [17 62,91,21,31,97,57,55,40,52,60,52,73,46,63,46,28,72,10,36,28,74,100,65,11]; // Duration of tasks

float goal=184; //Average Duration of Tasks assigned to each employee (Total duration of all Tasks/Number of Employees)
 OPL Code

//Objective Function
//Minimize the sum of Upper side and Lower side deviation in assigned tasks for each employee from the goal
minimize sum(j in N) (d1[j]+d2[j]);
//Constraints
subject to {
// Constraint that each task must be assigned to exactly one employee (cannot split among employees)
C1 : forall (i in M)

sum(j in N) x[j][i] == 1;
// Constraint that each employee will handle a minimum of one task (No one with zero task assignment)
C2 : forall (j in N)

sum(i in M) x[j][i] >= 1;

 OPL Code

// Constraint on deviation variables for each employee's goal

C3 : forall (j in N)
sum(i in M) (dr[i] * x[j][i])- d1[j]+d2[j]==goal;
// Constraint on Non-negative overhead deviation variable
C4 : forall (j in N)

d1[j] >= 0;

// Constraint on Non-negative underhead deviation variable

C5 : forall (j in N)

d2[j] >= 0;
}
Duration based Fair Scheduling of Employees
Employees Tasks assigned Duration assigned Deviation

Employee 1 1, 3, 9, 20 [ 17 + 91 + 40 + 36] = 184 0

Employee 2 11, 12, 18 [ 60 + 52 + 72] = 184 0

Employee 3 13, 23, 25 [73 + 100 + 11] = 184 0

Employee 4 4, 8,10, 16,19 [21 + 55 + 52 + 46 + 10] = 184 0

Employee 5 14,15, 22 [46 + 63 + 74] = 183 1

Employee 6 5, 6, 17, 21 [31 + 97 + 28 + 28] =184 0

Employee 7 2, 7, 24 [62 + 57 + 65] =184 0