Theory of Constraints in Software Engineering
Developed by Iðunn Tara Ásgrímsdóttir
(TOC) is a methodology invented by Dr. Eliyahu M. Goldratt a scientist, physicist, author, educator and consultant. With the methodology the most important limiting factor, that hinders a goal to be achieved, is identified. Then that factor (i.e. constraint) is improved until it is no longer the limiting factor. Since Goldratt introduced the Theory of Constraints in his bestselling 1984 novel, "The Goal", the methodology has continued to evolve and develop into many different fields including Software Engineering. 
The goals of Software Engineering is the creation of software systems that meet the needs of customers. The systems should be efficient, maintainable and reliable in addition to meeting project schedules and budgets and be produced in an economical way. Applying analogous set of approaches to the development of software, one can expect the production to be a software that meets the goals of Software Engineering. 
Theory of Constraints (TOC) in relevance to Software Engineering is two folded. First of all, it is used to improve the flow of new products to the market and second determining the real value of a proposed project or a feature to the final user. The assumption is that if the real value to the user is known, then it is possible to develop the right marketing and sales approach to materialize the value to the user and the software organization. 
This article will introduce how the Theory of Constraints can be used in Software Engineering to achieve a set of previously identified goals (projects for example). The article will also give guidance on how to apply and use the methodology in practice.
Theory of Constraints
Before improving a system or any part of it, one must define the system’s global goal and the measurements that will be used to evaluate the effects of any subsystem or local decision on the goal. When measurements have been defined, the constraint that limits the system to reach its goal can be found. 
Theory of Constraints (TOC) is a methodology that is used to identify the most important constraints in a system and then improving the constraint until it is no longer the limiting one. According to TOC every complex system consists of multiple linked activities and one of these activities can be a constraint upon the whole system.
Constraints are like said before, anything that prevents the organization from making progress towards its goal. Constraints can take many forms and are not necessarily only equipment. There are differing opinions on how to categorize different constraints, them, one approach is shown in the table below:
|Physical||Typically equipment, but can also be other tangible items, such as material shortages, lack of people, or lack of space.|
|Policy||Required or recommended ways of working. May be informal (e.g. described to new employees as “how things are done here”). Examples include company procedures (e.g. how lot sizes are calculated, bonus plans, overtime policy), union contracts (e.g. a contract that prohibits cross-training), or government regulations (e.g. mandated breaks).|
|Paradigm||Deeply ingrained beliefs or habits. For example, the belief that “we must always keep our equipment running to lower the manufacturing cost per piece”. A close relative of the policy constraint.|
|Market||Occurs when production capacity exceeds sales (the external marketplace is constraining throughput). If there is an effective ongoing application of the Theory of Constraints, eventually the constraint is likely to move to the marketplace.|
Theory of Constraints provides powerful set of tools to achieve the ultimate goal of a company: Make a profit .
Two of those tools will be introduced in the following chapter and later on explained in relation to Software Engineering along with explanation on how to use them.
The Five Focusing Steps
TOC provides the Five Focusing Steps to identify and eliminate the constraints, the steps are described in the table below:
|Identify||Identify the current constraint (the single part of the process that limits the rate at which the goal is achieved).|
|Exploit||Make quick improvements to the throughput of the constraint using existing resources (i.e. make the most of what you have).|
|Subordinate||Review all other activities in the process to ensure that they are aligned with and truly support the needs of the constraint.|
|Elevate||If the constraint still exists (i.e. it has not moved), consider what further actions can be taken to eliminate it from being the constraint.
Normally, actions are continued at this step until the constraint has been “broken” (until it has moved somewhere else). In some cases, capital investment may be required.
|Repeat||The Five Focusing Steps are a continuous improvement cycle. Therefore, once a constraint is resolved the next constraint should immediately be addressed.
This step is a reminder to never become complacent – aggressively improve the current constraint… and then immediately move on to the next constraint.
Throughput Accounting is TOC´s approach to accounting, a decision tool designed to ensure decisions that might have a financial impact take a holistic perspective and do not optimize one metric or area of the business at the expense of the whole system. Throughput Accounting takes a decision regarding the constraint and values it according to its financial terms. It acknowledges that the constraint determines capacity, hence profit potential of the company. Cost accounting is known to perform gross margin analysis at the product level while Throughput Accounting determines profitability at the system level and looks at the production process as a single system that must be optimized. Irrespective of the number of units created, Throughput Accounting assumes that most production costs are required to maintain a system of production. The approach also assumes that most production costs do not vary directly with incremental production of a single product.
Theory of Constraints in Software Engineering
Theory of Constraints was first introduced in relation to manufacturing. In manufacturing the goal is to identify bottlenecks in the production line. The constraint can be found where there are queues of Work In Progress (WIP) in front of specific equipment and therefore causes extra inventories.
Since the method was first applied, has it been applied in project management, engineering, sales, accounting and other business processes. In Software Engineering, the whole idea is to identify and manage what causes the inventory, but what is inventory in Software Engineering?
The list of client-value functionality that Software Developers have to solve is an inventory in Software Engineering. 
Inventory in Software Engineering can be expressed in different ways, depending on the software methodology used. A unit of inventory could be, for instance:
- A Use Case in UDP [a.k.a. RUP].
- A Story Point in Xp (eXtreme programming) .
- A Feature in FDD (Feature-Driven Development).
- A Backlog Item in Scrum.
- A Function Point in traditional SDLC structured methods. 
Applying Theory of Constraints in Software Engineering
According to David J. Anderson, producing and holding too much inventory is really bad in software because of the perishable nature of the inventory where requirements can go stale because of changes in the market or the fickle nature of the customer. Requirements have time value. With time, the throughput from the transformation of the requirement into working code decreases, specially when requirements are replaced by change requests for new requirements with a current market value. Requirements that become stale are pure waste .
In this chapter the applying of the theory will be explained through Five Focusing Steps and Throughput Accounting in relevance to Software Engineering
Five Focusing Steps
1. Identify the system´s constraints
The first thing to do in order to use Theory of Constraints to improve a system is to identify the constraints and prioritize them according to their impact on the goal. As said before, constraints can take many forms and the most known constraints are bottlenecks and therefore it can be a good idea to start identifying them. It is generally assumed that there is only one global system constraint at any given time. Once a constraint has been identified, a decision must be made on how to minimize its constraining ability on the system. The constraint must be fully utilized and according to David J. Anderson; every unit of production lost in the constraint is a unit lost in the whole system .
2. Decide how to exploit the system´s constraints
When the constraints have been identified and prioritized, a decision about how to manage them has to be made. The vast majority of the system´s resources that are not constraints have to be managed also.  The objective off this step is to make the most of what you have and maximize throughput of the constraint using currently available resources. The step focuses on quick wins and rapid relief; leaving more complex and substantive changes for later. 
In the following table items that could improve constraint are described in general:
|Buffer||Create a suitably sized inventory buffer immediately in front of the constraint to ensure that it can keep operating even if an upstream process stops.|
|Quality||Check quality immediately before the constraint so only known good parts are processed by the constraint.|
|Continuous Operation||Ensure that the constraint is continuously scheduled for operation (e.g. operate the constraint during breaks, approve overtime, schedule fewer changeovers, cross-train employees to ensure there are always skilled employees available for operating the constraint).|
|Maintenance||Move routine maintenance activities outside of constraint production time (e.g. during changeovers).|
|Offload (Internal)||Offload some constraint work to other machines. Even if they are less efficient, the improved system throughput is likely to improve overall profitability.|
|Offload (External)||Offload some work to other companies. This should be a last resort if other techniques are not sufficient to relieve the constraint.|
David J. Anderson suggested to use exploit Software Developers, that work 8 hours per day, as resources in the following ways:
- Always have pool of development tasks ready to be done for the developer
- Protect the developer from interruptions by providing a communication structure that minimizes the lines of communication
- Provide quiet environment for working to protect the developer from distraction
- Provide the best Software Development tools available to further exploit the developer
- Provide support staff for nonproductive activities (progress reporting, time-tracking and more) - Non value-added work is waste
- Provide adequate training in the technologies being used
- Provide team of colleagues to support and help resolve difficulties
- Requirements she is given are of good quality 
If the actions taken in this step solve the constraint the next step is to go on to step five, otherwise continue to step three.
3. Subordinate everything else to the above decision
The primary objective in step three is to support the needs of the constraint as the focus is on non-constraint equipment .
Some useful techniques for this step include:
|DBR||Implement DBR (Drum-Buffer-Rope) on the constraint as a way of synchronizing the manufacturing process to the needs of the constraint.|
|Priority||Subordinate maintenance to the constraint by ensuring that the constraint is always the highest priority for maintenance calls.|
|Sprint||Add sprint capacity to non-constraint equipment to ensure that interruptions to their operation (e.g. breakdowns or material changes) can quickly be offset by faster operation and additional output.|
|Steady Operation||Operate non-constraint equipment at a steady pace to minimize stops. Frequent inertial changes (i.e. stops and speed changes) can increase wear and result in breakdowns.|
To fully exploit constraints in Software Engineering the developers working should not be loaded with any more than they can reasonably handle. Thus the rest of the system of the software production must be subordinated to this notion.
4. Elevate the system´s constraints
In step 4 the objective is to elevate the constraint and might thus require more substantive changes. Investment of time and/or money is something that might be needed to elevate the constraint. .
Elevating a software developer from a constraint can be done in various ways. According to David J. Anderson the most obvious way used in the software industry is unpaid overtime, therefore the constraint in stretched. Managers could also hire more developers, introduce shift working or even use a better developer .
5. If in the previous steps a constraint has been broken, go back to step one, but do not allow inertia to cause a system constrain
Throughput Accounting (TA)
In traditional accounting, inventory is an asset that often drives undesirable behavior at companies to manufacture items that are not truly needed. Those items are manufactured because it generates a "paper profit" based on inventory that may or may not ever be sold. Since the Theory of Constraints considers inventory to be a liability, Throughput Accounting is used as it attempts to eliminate harmful distortions introduced in traditional accounting practices.
Core measures in Throughput Accounting are the following
|Throughput||T = Revenue - Totally Variable Expenses|
|Net Profit||NP = Throughput - Operating Expenses|
|Return on Investment||ROI = Net Profit / Investment|
According to Steve Tendon, to make the correct decision, you need a positive answer to one of these three questions:
- Does it increase throughput?
- Does it reduce operating expenses?
- Does it increase the return on investment? 
Positive business decision can be taken if the action considered increases throughput, decreases operating expenses or increases return on investment. 
Throughput Accounting for Software Engineering has been defined the following way:
- Throughput: is the rate of cash generated through delivery of working code into production […] It is computed as sales price minus direct costs, such as packaging, delivery, installation, training, support, and networking.
- Investment: is all money invested in software production systems plus money spent to obtain ideas for client-valued functionality. It thus includes software development tools and requirements gathering. […]
- Operating Expense: is all money spent to produce working code from ideas. It is primarily direct labor of software engineers, but it also includes selling, general, and administrative costs. 
Three examples will be given here below to apply Throughput Accounting in Software Engineering:
- Reducing Operating Expenses can be done by discarding requirements before starting implementation work. Increasing throughput is done by eliminating user stories in an inventory that are worth less, therefore Operating Expenses have been reduced.
- Investments can be decreased by operating an Open-Source software. The more the Open - Source solution covers of the requirements of a company the more decrease in Investment.
- Targeting on a market niche instead of a broad market, a unit price can be increased and therefore the throughput is increased.
When applying the Theory of Constraints, one has to be aware of its limitations before implementing it. First of all, the identification of the constraint is really important. The theory might work on a constraint that is caused by another constraining factor or it might focus on an irrelevant constraint. Focusing on irrelevant constraint could lead to waste of resources and time to limit some problem that is not critical to the success of the company. Another limitation of the theory is lack of consideration of variable factors. If the constraint is a demand for product, that varies independently from any action taken through the theory, resources invested in increasing the demand could have been more beneficial in expanding the production capacity. Theory of Constraints limits itself to short-term effects which is a limitation that can be overcome by examining the long-term effects of the work. If the short-term effects remain valid over a longer time-frame, the strategy indicated by the theory may be valid. If not, then other constraints have to be identified that result in long term benefits.
-  Vorne, Theory of Constraints This guide describes Theory of Constraints in comprehensible terms. It focuses on the basic terms of the theory and gets right to the point in each matter. The guide is made by an organization that specializes in manufacturing systems and follows the latest trends in Lean Manufacturing. The guide contains helpful tables about the subject which are used through the article. The guide is used through-out the article to get a better and simpler understanding of TOC.
-  Eric J. Braude and Michael E. Bernstein, 2016, Software Engineering Modern Approaches. This book is designed to accommodate multiple approaches to the learning and teaching of Software Engineering. It is good to understand the basic thinking behind Software Engineering. The book was used in the article to get explain Software Engineering and its concept.
-  David J. Anderson, 2003, Agile Management For Software Engineering: Applying the Theory of Constraints for Business Results This book is about software engineering management and explains how agile software development is used to make a business better. The author shows managers how to apply management science to gain the full business benefits of agility through application of the Theory of Constraints.
-  Eliyahu M. Goldratt, 1990, What is this thing called Theory of Constraints and how should it be implemented? This book walks through the crucial stages of a continuous program: the five steps of focusing; the process of change; how to prove effect-cause-effect; and how to invent simple solutions to complex problems. Goldratt is the one who originally introduced TOC and his way of introducing it is a little bit different than the others. The book is used in this article to get a better understanding of the Five Forces and TOC in whole.
-  Steve Tendon, 2012, Theory of Constraints and Software Engineering. This article is looking at TOC and how it can be applied to Software Engineering Management. The article examines how Throughput Accounting can be used to take important management decisions. The article compares Cost Accounting versus Throughput Accounting with Software Engineering in mind along with explaining Throughput Accounting for Software Engineering. The article is used to get a better understanding on how Throughput Accounting is used in Software Engineering.
- ↑ McMullen, Jr., Thomas B. "Introduction to the Theory of Constraints (TOC) Management System", APICS Pub.1998
- ↑ 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 Unknown. "Theory of Constraints", Lean Production, Date unknown. Retrieved on 10 September 2016.
- ↑ 3.0 3.1 3.2 Braude, Eric J. & Bernstein, Michael E. "SOFTWARE ENGINEERING Modern Approaches", Waveland press, INCUnited States of America 2016.
- ↑ 4.0 4.1 4.2 4.3 4.4 Anderson, David J. "AGILE MANAGEMENT FOR SOFTWARE ENGINEERING: Applying the Theory of Constraints for Business Results", Prentice Hall PTR United States of America, September 17 2003. Retrieved on 10 September 2016
- ↑ 5.0 5.1 5.2 Goldratt, Eliyahu M. "What is this thing called THEORY OF CONSTRAINTS and how should it be implemented?", ', Croton-on-Hudson, North River, New York 1990
- ↑ Tarantino, Anthony & Cernauskas, Deborah "Risk Management in Finance: Six Sigma and Other Next-Generation Techniques", John Wiley & Sons, Inc, 2009, chapter 21
- ↑ 7.0 7.1 7.2 7.3 7.4 7.5 7.6 7.7 Tendon, Steve. "Theory of Constraints and Software Engineering", The TameFlow Chronologist, Published July 27th 2012. Retrieved on 13 September 2016.
- ↑ Bert Markgraf. "Limitations of the Theory of Constraints", studioD, Date unknown. Retrieved on 25. September 2016.