|A unit of Lasting Forests
evolving since March 30, 1999
A Total Forest Management Plan
It's hard to see the future with tears in your eyes.
Mohawk saying reported by Wilma Mankiller,
former chief of Cherokee Nation
Feedforward does not seem to be a clearcut topic. We think it is an important concept and so we make two small presentations ...
Plans are about the future. A plan may exist for the present, as in "I have a plan", but it addresses the future. One element of superior plans is feedforward. A rarely used word, it is best understood by comparing it to feedback. The best popular example of feedback is the thermostat in the house. It is a procedure that senses the temperature, compares it to the desired temperature, and holds-fast or changes the furnace or air conditioner functions until the desired temperature is reached. It is one named "thing", but it is composed of several actions. Feedforward has parallels to feedback. Feedforward predicts the future, compares it to objectives, and makes adjustments in the major elements of the present system so that the total system will be very responsive over time to the predicted future conditions.
If a forester predicted rapidly growing interest in a special type or color of wood, he or she might re-forest appropriate areas with those species of trees. In effect, they would be getting ready for the predicted future. Feedforward is not the prediction itself (called "a prediction", a "forecast", or a "future") but the total process (a mini-system) of preparing now for the future.
Another example may be useful. A building may be needed and the number of people to be served is estimated to be 200. The costs may be too great to build for 200 and no more than 100 would be served in the first 5 years. A building is then built for 100, full well knowing it is too large now (there are only 40-50 potential users), just right for 100 later, but too small for the growing group. Feedforward involves taking risks and careful analysis of the meaning of being "most right over the longrun." In some cases average annual performance will be a good measure. In other cases, there is a need to keep the deviations (e.g., peak loads) small.
Another example is very important for some agencies and programs. A plan is written. It is reasonable to write several scenarios for the future if certain laws or policies (now in the "hopper" ) are passed. We know future conditions will be different; we even know that the change agent is active. We must put in the present plan the responses to the several limited likely futures. This is a specific act of feedforward.
No one can know the future. Feedforward was omitted from early development of general systems theory for it was unknown in biological systems (only in humans). The future can be estimated (estimates about the future are called "predictions"). Some things are known at such a high probability that they can be assumed to be known (e.g., water running downhill, the sun rising). Otherwise it seems reasonable to discuss the probability of future events.
Some psychologists believe that some people never reach a stage of mental development or maturity in which they can see or imagine the future. Few people buy insurance, implying some things about their expectations, concept of the future, and what, if anything, can be done now to prepare for that likely state. Few people have a "nest egg", partially because of economic conditions, partially because the likely future conditions, some pretty bad, cannot be very real to them.
We may not know the future price of wood, but we know about wood, prices in the past, and can see the influences of transportation, national economy, and housing starts on the price of wood. Our estimates of the future price are made with some confidence. Our accuracy is pretty good. We do know much about the future ...at least we can make good predictions. Knowing the price (or similar measurable units) can suggest things to do now to get ready (e.g., cut backs if there is an unlikely payoff; increase production if the costs are in line with net gains).
Foresters are among a special group of people who must use feedforward in society. Predicting the future and planting trees for that future is one of the most long-term projects with payoffs far outside of the limits for most current businesses.
There are methods and procedures for improving feedforward:
An example of a change analysis is that related to a sharply decreased supply of nitrogen; thus more areas in farms are needed to produce the same amount of food, thus less forests and wetlands, thus more intensive use of some forested areas, thus less intensive use of some areas (e.g., old growth).
A similar pattern might be used with decreases in the supply of phosphorus, thus major decreases in crops a, b, and c; thus reduced amounts in ponds, thus reduced eutrophication, also reduced crop wastes due to shift in production of crops ...and wildlife effects.
Feedforward includes the special concept of a failsafe action or operation. Depending on how sensitive to system is to failure or how harmful will be the consequences of failure, backup or failsafe mechanisms are needed. The failsafe action looks ahead, attempts to describe worst-case scenarios, predicts the consequences, and then attempts to take action at reasonable costs that will prevent a failure. An police person or agent traveling with another agent as "backup" is a simple example. A modern agency failing to get a license increase will likely have a failsafe mechanism. Endangered species, for example, may need intensive protection by officers, but that may not work or be insufficient or may contain lapse periods of no protection. A protective fence may be another simple example of a failsafe mechanism for this problem. An officer carrying two guns and an educator carrying 2 projectors are examples. For truly important events, double backup is always suggested (and even that fails, but the costs can rarely be justified, given the odds of multiple failures). Practical people on tight budgets will rarely tolerate having around apparently un-used things, the backups for the things in use.
Rabow (1969:38) observed that predictions as part of feedback can influence behavior. He said, "when credible prediction about the behavior of an individual or group are published, or otherwise learned by the subject, this may cause him to alter his behavior. Published predictions of social phenomena must therefore take into account the effect of prediction on the individual involved. This feedback phenomenon, although it may complicate the prediction process, may actually make the prediction more reliable." Rene Dubos (pers. Comm) observed that dire predictions are made, hoping that change will occur, and thus that the prediction will be wrong.
These are general and introductory comments about feedforward. Others are available within www.LastingForest.com. The proposed actions for later within The Trevey are:
Another look at the topic ...
Feedforward is not the easiest concept . We do not know why. Let us try to present it ... again. People who work with computers understand feedback, the corrective, adjusting, adaptive work that goes on to analyze inputs, programs, and the scale of the system (the context) to make it consistent with the objectives.
The room thermostat may suggest useful comparisons to understand feedback, then to compare it with feedforward. Similar parallels are evident for creating elk harvest systems or pollution abatement work in rivers. With feedback seen as a type of system itself (all of the components of general systems theory can be seen to be present), it seems reasonable to ask for some parallel thought about feedforward.
Feedforward is that subsystem that seeks to estimate or comprehend the future and then to revise the present system based on that perception. It is not just future-telling (any more than a thermostat is just a thermometer). It is not prognostics or futurism. It is a multi-technique approach to seeing the future and acting today to adjust the now-system to meet these tomorrow-system conditions.
Much needed, nevertheless some land owners are put off by the notion that feedforward makes today's system wrong or sub optimum but allows the system to be most-right or least bad, optimal, over the long run (however defined or decided).
A trivial example: a hunt club wants a meeting place. They can build it for the 10 members today. They discuss and estimate the future and imagine 20 members. The price for such a building is too great. They build for 12 people with an expensive window that can be converted into a door later. The meeting place is too large for today, too small for the longrun, but a reasonable design based on the perceived future. Feedforward, like feedback, is an action concept. Something happens when it is used.
A natural resource agency predicting that the number of hunters will decline drastically will not be adding permanent staff to teach hunter safety or creating new programs to produce more game for hunters.
Programmers often use feedforward. Building a program with a view for the future and with a means to add easily a new module or with an extra translation program to unify two programs is using feedforward. The present program is bigger and better documented than it now needs to be, thus inefficient. Over the longrun it may be very efficient. Estimating the future and taking action now to respond to that future is a very human activity - nowhere else present in nature.
Go to top of page.
|Quick Access to the Contents of LastingForests.com|
This Web site is maintained by R. H.
Last revision August 25, 2004.