Decision analysis as a basis for more effective agricultural innovation
Part 3: Tacit knowledge & performance
The most important factors in project design and implementation performance are the ability to enhance the capabilities of team members, clarify their task assignments and to provide support for their work. Individual capability depends upon qualifications in terms of training, experience and the ability to apply know how effectively as well as to learn on the job. The acquisition of capabilities depends upon two forms of knowledge:
Explicit knowledge is that information that can be communicated by word of mouth, in written form, sound, imagery and musical notations or sounds which today can be distributed, stored, accessed and processed through digital technology. The main structure to this information is language, mathematical logic, such as Boolean logic and verbal argument. Mathematical formulae are a useful component of explicit knowledge as are ways and means of building decision analysis models and applying analytical techniques and simulation. In summary, explicit knowledge can be passed on from one person to another and from generation to generation in the form of records and it is the medium used in educational and training activities.
Tacit knowledge is an operational capability, or skill, associated with each individual in the carrying out of a task. Tacit knowledge is developed as a result of an individual becoming acquainted with the application of a technique and applying it on a repetitive basis. The difference in quality of output and efficiency in execution of an experienced practitioner over a novice is the difference in their levels of accumulated tacit knowledge. Tacit knowledge is usually related to the well-established phenomenon of the "learning curve". The effect of the learning curve was explained by Theordore Wright in a paper in 19362. The learning curve impact described by Wright has also been referred to as Wright's Law has turned out to be a reliable basis for predicting the impact of tacit knowledge on the productivity of organized processes involving a combination of technologies and humans.
Learning curve and index
The learning curve is the phenomenon of the occurrence of measurable reductions in the resources used (including time) in the production of an object in association with the cumulative quantity of throughput.
The heuristic, or rule of thumb, is that the time required or resources used to complete action fall by a constant percentage for every historic doubling of output. This means that the gain from learning on a production amount of x units will be the same as the gain from 2x units and this will be the same as 4x units and 8x units should have the same gain in performance.
This gain in performance, because of the historic doubling of throughput factor signifies that there are diminishing marginal returns which complies with production theory.
Geometry of the learning curve
The learning index is the percentage reduction in resources used with each historic doubling of throughput. This is also expressed as a percentage curve thus more capital intensive processes might have a 90% curve signifying 10% reductions and a more labour–intensive process might have an 80% curve indicating a 20% reduction. An example of a 80% curve is shown on the right.
Some general relationships influencing learning curves
Belkaoui4 cites a summary by Hirschmann5 of the basic doctrine of the learning curve as:
In general terms the learning curve effect is more pronounced if production processes are labour-intensive and less pronounced if production processes are more capital-intensive or automated
Throughput and quality of output
Throughput is measured in terms of quantity and quality. The acquisition of tacit knowledge through learning tends to be associated with higher quantities of output per unit time and improved quality of output. The quality of output is measured by the quantity of output that meets a specified quality. The yield of a process is the percentage of output that meets quality standards. Usually an individual will have a predefined maximum capacity for production (Cp) according to a the process equipment, tool sand applied technique. This is measured in terms of units of output per hour or day. The achievement of that maximum quantity of throughput depends upon the time allocated each day to the task (Ta) and the capacity utilization (Cu). Therefore the quantity of physical output is given by:
and the quantity of output of the required quality is given by:
The capacity utilization and yield are strongly correlated to tacit knowledge or skill in the carrying out of tasks. Capacity utilization can be related to the operator applying the correct equipment settings or following efficient procedures to take full advantage of the technique being applied. The quality yield is usually related to the level of skill applied to operations. The table below should typical relationships associated with experience of operators:
As can be observed unit costs and waste decline with learning while output quality and capacity utilization and yield increase with the level of tacit knowledge.
The scope and of relevance the learning curve to agricultural innovation projects.
The main characteristic of the learning curve is its reliance on repetitive application of a technique. It is therefore evident that the impact is greater the more extensive the time during which specific techniques are applied. Sometimes, within the space of a project there will not be sufficient experience built up to have any significant impact. However, innovation projects have an unusual context which involved several stages as illustrated in the diagram on the left. These stages include prioritized research, proof of concept, prototype creation and following feasibility studies, investment for production. The longest "run" or operational time would normally be associated with the investment in the form of the commercial production where learning curve advantages usually become more apparent. However, some of the projects leading up to that stage can become be altered to switch from being one-off projects to continuous processes as an integral part of the production chain supplying the commercial operation. Therefore, the relevance of learning curve projections will vary according to the innovation stages.
How Navatec System integrates learning curve projections
Navatec System make use of the learning curve relationships to provide projections of the impact of rises in human capability in the application of given technologies and techniques. This contribution to project performance is related to the use of resources and time to complete each operation. Navatec System has an embedded learning curve server side utility (SSU)w that enables raw performance data to be input to calculate the learning index and then project costs or timing. If benchmark data is available this can be input to obtain the same results.
Reviewing impact of learning using Monte Carlo Simulation
Beside generating accurate evidence-based projections for learning impact the Navatec System also provides the means to simulate different scenarios using Monte Carlo Simulation (MCS) to establish feasible targets for performance within given timeframes. The image below is a screen shot of one of the simulators input dialogs for the MCS that deals with quality and quantities of throughput.
Project design and appropriate training
Irrespective of any gains that might accrue from the build up of tacit knowledge, the maximisation of the contribution of techniques is related to the deployment of the most appropriate best practice in terms of efficiency, effectiveness and economy within the constraints facing a project.
What defines best practice is normally based on experience and benchmark data where the current practice has benefited from the contributions of tacit knowledge in refining the techniques. Thus within each technique there is a constant innovation in how it is applied and in the attainable performance.
Usually the benchmark performance will accord with the current levels of experience of team members in applying the technique in question. Training has an important role in making sure that:
Naturally, each team member needs to descend the learning curve and they cannot start applying a technique with the competence of a skilled and experienced person. However, with an adequate knowledge of performance benchmarks that relate to learning curve coefficients, it is possible to forecast the rate of increase in performance associated with experience in applying the technique. With use of a technique in a work environment the person concerned will build up an appreciation of a sequence of actions and their relationship to the quantity and quality of output achieved. Quality of output, which is measured by applying explicit knowledge will become clearly related to the way in which specific manipulations are carried out so the tendency to refine those actions result in the more experienced person producing higher quality output. Cumulative tacit knowledge has an important role in raising the performance of activities.
1 McNeill, H. W., "The State-of-the-Art & Future of Decision Analysis", SEEL, The George Boole Foundation, BSI,HPC, October, 2009
2 Wright, T., "Factors Affecting the Cost of Airplanes", Journal of Aeronautical Science, Volume 3, No.2, 1936, pp. 122-128.