TRIZ and Software Fini
trizjournal | On 10, Aug 2004
By: Ron Fulbright
Director, Information Management & Systems
University of South Carolina Spartanburg
Spartanburg, SC 29302
In  and , Rea discusses software engineering analogs to Altshuller‚Äôs 40 principles of innovation. However, Rea did not list analogs for principles: 29 (pneumatic/hydraulic construction), 31 (porous materials), 36 (phase transition), 37 (thermal expansion), 38 (accelerated oxidation), and 39 (inert environment). This article completes Rea‚Äôs list by offering software analogs for these six principles and summarizes all 40 software TRIZ principles.
2 Additional Software Analogs
2.1 Hydraulic/Pneumatic Construction (Principle 29)
For mechanical systems, this principle applies to variable-volume parts inflated with liquid or gas. In software, the analog of a container with varying volume is a dynamically allocated data structure. It is often unknown at compile time how much data a program will need to handle. With dynamically allocated data structures, as more elements are required, the data structure‚Äôs memory footprint expands and then contracts as elements are deleted.
2.2 Porous Materials (Principle 31)
We like to think that our software produces the correct output at all times. However, sometimes this is not always desirable. The suggestion here is that ‚Äúporous material‚ÄĚ be interpreted as intentionally making a software application imperfect. An example is an intelligent tutoring system (ITS). Consider a student learning to play chess from an ITS. The student will become frustrated and less likely to enjoy playing if the computer wins all the time. Also, if the computer plays perfectly all the time, the student will not learn to take advantage of opponents‚Äô mistakes‚ÄĒa critical skill in playing chess with human players. Therefore, the ITS needs to be ‚Äúporous‚ÄĚ and intentionally make mistakes to play down to the level of the student. Indeed, the degree to which the ITS does this is can change over time and in concert with monitoring the student‚Äôs progress via another TRIZ principle, feedback.
2.3 Phase Transition (Principle 36)
Recent research in nonlinear dynamics has exposed an interesting feature of complex adaptive systems. It seems that in a dynamical region just on the controllable side of chaos is a regime called the emergent regime in which systems achieve the highest levels of global emergent behavior. Researchers in artificial life have explored this region and coined the phrase ‚Äúlife at the edge of chaos‚ÄĚ to describe the sudden onset of complex and sustainable patterns in that region. Others have applied the same idea to natural complex adaptive systems like biology, economics, and markets. Wolfram envisions using the phenomenon as a whole new approach to science.
Researchers liken the sudden shift of a system from controlled behavior to emergent behavior to the change of phase in physical systems‚ÄĒlike water changing from solid to liquid as it melts. The degree of randomness in these systems is a key parameter. It seems that given the right amount of randomness, a complex system can be induced to change phase to the emergent regime in which its information processing capability is maximized thereby allowing the system as a whole to achieve more than the sum of its parts. This certainly applies to software systems.
2.4 Thermal Expansion (Principle 37)
Thermal expansion or contraction in physical systems involves a volume change as an object is heated or cooled. A computer‚Äôs memory space is a combination of active memory (in the CPU) and paged memory (maintained in some nearby storage medium such as cache or virtual memory). The expansion and contraction of this resource, in response to more or less processes requiring varying amounts of memory can be modeled thermodynamically by attaching a metric analogous to temperature to the system which would then model the computer‚Äôs performance at a given time.
2.5 Accelerated Oxidation (Principle 38)
In chemical systems, oxidation is the process of combining with oxygen thereby releasing energy stored in the chemical bonds. This reaction produces heat, a randomized quantity of energy. Obviously, software does not bind with oxygen, but we can abstract the oxidation principle to refer generically to the mixing of something with something else to produce a randomized output. Salted encryption comes to mind as an analog. An encryption algorithm without a random component, ‚Äúsalt‚ÄĚ, run on some cleartext (say user passwords) will always produce the same encrypted output. A particular password would always be encrypted to the same string on every computer running the unsalted encryption algorithm. If you crack the password once, you can evade security on every other computer employing that algorithm. However, if the encryption algorithm adds a random factor, called ‚Äúsalt‚ÄĚ, into its calculations, the encrypted text is valid for only the one machine, since, theoretically, all other machines would salt their calculations differently.
2.6 Inert Environment (Principle 39)
An inert environment is one that tends to not react with objects in the environment. A logical connotation is that an inert environment is a benign one. With this interpretation, software test harnesses serve as an analog. In software development, it is often necessary to test the software being developed in a simulated environment providing some, but not all, of the behavior of the actual environment the software will operate in. This artificial construct is generally called a ‚Äútest harness.‚ÄĚ
Another analog is benchmark tests, often used to measure hardware and software performance. The environment in which the benchmark is run is carefully controlled to insulate the system from uncontrolled influences while retaining critical characteristics and thus is also an inert environment.
3 Summary of TRIZ Software Analogs
Combining the above analogs to Rea‚Äôs analogs, and editing for space, results in the condensed summary of TRIZ analogs for software shown in Table 1.
 Rea, K.C., TRIZ and Software 40 Principles Analogies, Part 1. The TRIZ Journal.
Sep, 2001. Internet: http://www.triz-journal.com/ archives/2001/09/e/index.htm
 Rea, K.C., TRIZ and Software 40 Principles Analogies, Part 2. The TRIZ Journal.
Nov, 2001. Internet: http://www.triz-journal.com/archives/2001/11/e/