8. LP Modeling: Production Planning

Let us look at another example of modeling a problem using Linear Programming, again to do with production.

So, recall that a linear program is an optimization problem, where you have some variables which describe the quantities you want to compute. And you now have linear constraints on these variables, as well as a linear function describing what it is that you want to optimize, maximize or minimize, that is called the objective function.

This chunk introduces linear programming as a method for solving optimization problems. It defines how linear programming involves variables representing quantities along with linear constraints and an objective function that can either be maximized or minimized. Essentially, when you're working with a linear program, you're trying to find the best possible outcome under a series of restrictions.

Think of a farmer who wants to maximize the yield of crops on limited land. The farmer has specific amounts of labor and resources, like water and fertilizer (these are the variables), and must follow rules about how much of each crop can be planted (the constraints). The overall goal is to find the best mix of crops that gives the highest yield (the objective function).

One way to solve a linear program is to think of it geometrically and uses simplex algorithm and the simplex algorithm exploits the fact that the optimum value of a linear program is always at the vertex of the fusible region.

The simplex algorithm is a method used to solve linear programming problems efficiently. It operates on the principle that solutions to the optimization problem are found at the vertices of feasible regions defined by the constraints. This means that instead of testing every possible combination of variables, the algorithm only needs to evaluate these corner points, making it much faster.

Imagine you are exploring a mountain area looking for the highest peak, but you can only walk along the edges of the trails (the vertices of the feasible region). You would start at one peak and move to neighboring peaks, comparing their heights until you find the highest one, instead of wandering randomly all over the area.

The next example we are going to look at is the carpet manufacturing company. So, we have a company which makes hand woven carpets and we currently employee 30 employees, each employee produces 20 carpets a month...

This chunk introduces a real-world example to illustrate linear programming. The carpet company employs 30 workers, each producing carpets at a fixed rate, but the demand fluctuates throughout the year. The company must plan accordingly to avoid excess costs due to unsold stock or insufficient supply during peak demand periods.

If you own a bakery, you need to predict how many pastries to bake each day based on past sales data. If you bake too many, you waste ingredients and money. If you bake too few, customers might leave empty-handed. Just like the carpet example, it requires planning based on expected demand.

Now I have to come up with this strategy to cope with this varying demand, in order to make sure that I lose the least amount of money, because of this coping demand fluctuation...

This section discusses strategies for dealing with variable customer demand. Options include paying workers for overtime, hiring additional employees, or storing surplus production. Each strategy has associated costs that must be considered when planning production to ensure profitability.

Imagine planning a party where you expect different numbers of guests. You might order extra food to prepare for a larger crowd (increasing cost), hire a caterer versus cooking yourself, or try to predict how much food to make based on past party sizes. You want to avoid running out of food while not overspending either.

So, we want to make a linear program out of this, so we need some variables. So, we are dealing with these 12 months January, February to December. For month i where i ranges from 1 to 12, let us assume that we have Wi workers...

This part details how to create a linear program by defining variables that represent the number of workers, carpets produced, and other pertinent figures for each month of the year. These variables are essential for setting up the linear equations that will define the constraints.

In planning the party, you’d define variables like the number of guests, food items needed, and staff required. For each week or month leading up to the event, you would need realistic estimates to ensure everything goes smoothly.

So, now let us think of the constraints, so the first constraint to the usual one, which is their every quantity that we are talking about is strictly greater than equal to 0...

This chunk outlines the constraints that must be included in the linear programming model. These include ensuring that quantities produced and stored cannot be negative and that the resources used do not exceed available amounts. Constraints are necessary to model the limitations the company faces realistically.

If you are baking, a constraint might be the maximum amount of flour you can purchase. You can't make pastries with a negative amount of flour, similar to how the carpet company can't produce negative carpets. Constraints guide how realistically you can fulfill your plan.

So, now given these constraints what we want to do is minimize the total amount of cost that we are going to put up...

This section discusses the objective of the linear program: to minimize overall costs incurred by the carpet manufacturing process. It categorizes the different types of costs involved including labor, hiring, firing, storage, and overtime, illustrating how these will be combined into the objective function.

Just like organizing a big event where you need to budget for venue rental, catering, and decorations, the carpet company's aim is to spend the minimum amount necessary while meeting production needs and fulfilling demand.

So, we done simplex and we find the solution, so are we done, so turns out that we might with an answer which is not something that we can actually used...

This final chunk highlights a potential issue with linear programming – that the solutions discovered through the simplex algorithm might involve non-integer values where integer solutions are needed (like hiring a fraction of a worker). It introduces the concept of rounding to create practical integer solutions but cautions that this can impact cost effectiveness.

Going back to the bakery scenario, if you need to hire staff for peak times, you can't hire half a person. You would need to round to the nearest whole number of staff, which could change your budgeting or lead to understaffing at certain times.

--