Main menu


Maintenance Implications on Cost and Schedule

Software Maintenance Implications on Cost and Schedule


 One of the finest demanding situations facing software program engineers is the control of alternate manage.

 It has been envisioned that the cost of alternate manage may be among 40% and 70% of the existence cycle expenses . 

Software engineers have hoped that new languages and new system would substantially reduce those numbers; however this has not been the case.

 Fundamentally this is because software is still added with a tremendous wide variety of defects.

 Capers Jones estimates that there are about 5 bugs according to Function Point created during Development . Watts Humphrey found "... Even skilled software program engineers typically inject a hundred or extra defects per KSLOC .

 Capers Jones says, "A collection of studies the disorder density of software program stages from forty nine Five to 94.

Five mistakes per thousand strains of code ." The reason of this newsletter is to first assessment the fundamentals of software maintenance and to offer alternative tactics to estimating software upkeep.

 A key element to be aware is that development and control choices made during the improvement technique can extensively affect the developmental price and the ensuing preservation fees.


 Maintenance activities consist of all paintings done publish-transport and should be outstanding from block changes which constitute tremendous design and improvement effort and supersede a formerly launched software program bundle.

 These upkeep sports can be quite numerous, and it facilitates to discover exactly what publish-delivery activities are to be protected in an estimate of renovation effort.

 Maintenance sports, once described, may be evaluated in a pretty distinctive light than while known as actually "maintenance". 

Software protection is different from hardware protection due to the fact software program does not bodily put on out, but software regularly gets much less useful with age and it could be brought with undiscovered flaws.

 In addition to the undiscovered flaws, it is commonplace that a few wide variety of regarded defects bypass from the development corporation to the maintenance organization.

 Accurate estimation of the attempt required to hold introduced software is aided through the decomposition of the general effort into the various sports that make up the entire method.


 Maintenance is a complicated and based procedure. In his textbook, Estimating Software Intensive Systems, Richard Stuzke outlines the typical software upkeep method. It is apparent that the procedure is greater than simply writing new code.

  • The following tick list may be used to explore the realism and accuracy of renovation necessities.
  •   Which pieces of software program might be maintained  ?

  •    How long will the system need to be maintained  ?

  •   Are you estimating the entire preservation trouble, or simply incremental upkeep  ?

  •   What level of upkeep is required

  •   Is that that's being called maintenance in fact a new improvement task ?

  •  Who will do the renovation ?
  •  Will or not it's done organically with the aid of the original  ?

  •  developer ? Will there be a separate group  ?

  •  Will there be a separate corporation ?
  •  Will maintainers be using the equal tools used at some stage in development ?
  • Are any proprietary gear required for protection ?
  •  How tons Commercial-Off-The-Shelf (COTS) is there ?
  •  How tightly coupled are the interfaces?

 Some follow-on development may be disguised as maintenance. This will either inflate upkeep figures, or else motive shortfalls if basic protection gets pushed aside.

 These questions will assist you ask whether protection is being genuinely represented.

  •  Is the interest honestly an incremental development?
  •   Are healthy chunks of the unique code being rewritten or modified?
  •   Will additional staff be brought in to perform the upgrade?
  •   Is the protection effort agenda regular and pretty flat, or does it incorporate ?staffing  humps that appear to be new improvement


Although sanity exams ought to be sought on a yr-with the aid of-year foundation, they must now not be tried for normal development. The reason for that is that preservation sports can be carried on indefinitely, rendering any life-cycle rules vain. As an instance,

We spend about 2 to 3 instances as an awful lot effort preserving and improving software as we spend creating new software program.

This and comparable observations practice at an organizational level and better, but not for a specific challenge. 

Any improvement organization with a history will be embroiled in the lengthy tail ends of their many added projects, still needing indefinite interest. Here are a few brief sanity tests:

  •  One maintainer can take care of approximately 10,000 strains in step with yr?
  •  Overall existence-cycle attempt is normally 40% improvement and 60% maintenance?
  •  Maintenance costs on common are one-6th of yearly development prices?
  •  Successful structures are commonly maintained for 10 to twenty years?

Finally, as in development, the quantity of code this is new as opposed to modified makes a distinction.

 The powerful length, that is, the equivalent effort if all of the work had been new code, remains the important thing input for each development and maintenance fee estimation.


 software estimation strategies must be capable of version the concept and the probably real international end result.

 The actual world scenario is that over the years, the overlay of adjustments upon adjustments makes software increasingly tough to hold and accordingly much less beneficial.

 Maintenance effort estimation techniques range from the simplistic stage of attempt method, thru more considerate evaluation and improvement exercise changes, to using parametric models which will use historic statistics to task future desires.

  •  Level of Effort A

 is from time to time the case inside the development surroundings, software program preservation may be modeled as a level of attempt hobby.

 Given the restore class activities and the extremely good variance that they display, this approach actually has deficiencies.

 In this approach, a degree of attempt to preserve software program is based on size and type.

  •  Level of Effort Plus

 Stuzke proposed that software program preservation starts with primary stage of effort (minimal humans had to have a core competency and then that that basic core body of workers have to be modified by using assessing three additional factors;

 configuration control, great warranty, and task management. His method addressed a number of the additional factors affecting software upkeep.

  •  Maintenance Change Factor

 Software Cost Estimation with COCOMO II (Boehm 2000) proposes a deceivingly simple, however additionally quite beneficial technique for figuring out annual renovation. Maintenance is one of the menu choices inside the menu bar.

 In COCOMO II Maintenance encompasses the system of enhancing existing operational software program whilst leaving its number one functions intact. 

: This manner excludes

- Major re-layout and re-improvement (more than 50% new code) of a new software product   appearing significantly the equal functions

- Design and improvement of a sizeable (greater than 20% of the source commands comprising the existing product) interfacing software program bundle which requires quite little remodeling of the existing product.

- Data processing device operations, facts entry, and change of values in the database

The preservation calculations are closely primarily based upon the Maintenance Change Factor (MCF) and the Maintenance Adjustment Factor (MAF).

 The MCF is similar to the Annual exchange Traffic in COCOMO81, besides that preservation intervals other than a year may be used. 

The ensuing protection effort estimation formula is the same as the COCOMO II Post Architecture improvement version.

As stated formerly, 3 value drivers for preservation differ from improvement. Those value drivers are software reliability, current programming practices, and agenda.

 COCOMO II assumes that elevated funding in software program reliability and use of modern programming practices during software development has a sturdy tremendous impact upon the upkeep stage.

Annual Maintenance Effort = (Annual Change Traffic) * (Original Software Development Effort)

The amount Original Software Development Effort refers to the total effort (individual-months or other unit of measure) expended during improvement, even if a multi-12 months undertaking.

The multiplier Annual Change Traffic is the share of the general software program to be modified in the course of the 12 months. This is fantastically clean to attain from engineering estimates.

 Developers regularly keep exchange lists, or have a sense of proportional change to be required even before improvement is complete.

  • Managing Software Maintenance Costs via Developmental Techniques and Management Decisions During Development

When it involves upkeep, "a penny spent is a pound saved." Better improvement practices (although extra high-priced) can significantly reduce maintenance attempt, and reduce standard lifestyles cycle price.

 The more attempt positioned into development, the less required in maintenance

 As an instance, the software improvement fee and agenda may be drastically impacted (reduced) by way of letting the number of defects brought grow. 

This value and agenda reduction is more than offset by way of the increase in maintenance price. 

The following discussion is an instance of ways control selection can extensively have an effect on/lessen software program renovation prices.

Lloyd Huff and George Novak of Lockheed Martin Aeronautics in their paper "Lockheed Martin Aeronautics Performance Based Software Sustainment for the F-35 Lightning II" propose a series of improvement and control decision designed to effect and reduce software program renovation costs.

 They propose an eight step manner to estimate and manage software program renovation . Their proposed steps are:

1. Strive for Commonality

2. Apply Industrial Engineering Practices to Software

3. Engage

4. Adopt a Holistic Approach to Sustainment

5. Develop Highly Maintainable Systems and Software

6. Manage the Off-the-Shelf Software

7. Plan for the Unexpected

8. Analyze and Refine the Software Sustainment Business Case (use Parametric software sustainment cost estimates)

  •  A Parametric Assessment of Software Maintenance

Parametric models like SEER for Software allow renovation to be modeled in both of two  approaches

Estimating upkeep as part of the overall lifecycle cost. Choosing the correct Maintenance class parameters will encompass an estimate of upkeep effort with the improvement estimate for the individual software program.

 Several reports and charts show breakdowns of improvement vs. Upkeep effort. This technique is best used to assess existence cycle fees for every individual software program software.

Estimating renovation as a separate interest. Using the proper maintenance parameters for the software program to be maintained you could version the upkeep effort as a separate interest.

 This technique will permit you to satisfactory song your maintenance estimate via adjusting parameters. Maintenance size have to be the same as improvement length, but must be entered as all pre-current code.

 This approach can also be useful in breaking out total mission upkeep prices from task improvement prices.

A exact parametric estimate for maintenance consists of a wide variety of records. Critical records for completing a software program upkeep estimate is the size or quantity of software in an effort to be maintained,

 the fine of that software program, the high-quality and availability of the documentation, and the kind or amount of protection in order to be achieved.

 Many companies don't surely estimate renovation expenses; they definitely have a budget for software renovation. In this case, a parametric model must be used to compute how lots protection can truly be performed with the given budget.

Estimating and planning for preservation are vital sports if the software is required to characteristic nicely at some stage in its predicted lifestyles.

 Even with a constrained finances, a plan may be made to use the assets available in the maximum green, effective manner.

 Looking on the diagram above, you could see that now not best are the a couple of inputs that effect the upkeep, but there are several key outputs that offer the information vital to devise a a success renovation attempt.

 Conclusion The conclusions of this article are

  •  Software renovation may be modeled the usage of a simplistic approach like Level of Effort Staffing, but this method has widespread drawbacks.
  •  Software renovation fees may be considerably affected by management selections during the developmental manner.
  •  Software upkeep may be correctly expected the use of parametric methods
  •  Software maintenance is great modeled when development and control decisions are coupled with parametric price estimation techniques.

table of contents title