Dependence on computers has intensified the division of labor within engineering

Question

Dependence on computers has intensified the division of labor within engineering. For example, civil engineers designing a flood control system have to rely on information and programs obtained from systems analysts and implemented by computer programmers. Suppose the systems analysts refuse to assume any moral or legal responsibility for the safety of the people affected by the flood control plans, arguing that they are merely providing tools whose use is entirely up to the engineers. Presumably their accountability does extend to errors resulting from their own inadequate specifications that they supply to the computer experts.

a. Should the civil engineers be held accountable for any harm caused by poor computer programs?

b. Should the engineers also be expected to contract with computer specialists who agree to be partially accountable for the end-use effects of their programs?

Explanation / Answer

Let me start by saying that this a very sensible and responsible question. What you've pointed out or rather observed is absolutely correct. There is a visible division of labour among engineering works.As in this case, Flood Control Tools when used produce results based on the input we have given.

First of all, one thing should be understood that such tools when developed undergo rigorous checks and often vetted by an eminent or experienced academician without fail, only after such rigorous checks, these programs are released out into the industry and academia. Also these tools are marketed on the basis of these checks i.e., analysis is done for the same case using manual means and by program. The variability of these results is taken as a key point in marketing. So large deviations of results from the theoretical or actual point of view is avoided to the maximum possible extent.

Now coming to the point of selection of a tool by civil engineer or any other engineer to that point. The most important step in selection of the tool is Calibration and Validation of the tool with existing data or analysis done by manual means. Any analysis carried out by tools that do not pass through these tests calibration and validation is on the head of the civil engineer as he/ she hasn't taken enough steps to ensure his results are satisfactory or are within the permissible limits. Also a good literature review of good journals may provide first hand information on the tool's merits and demerits, its versatility and applicability.

Let me explain with a very basic example of Flood Prediction of a catchment. Say you have flood data at a stream gauge point for 30 continuous years, say 1980 - 2010. You now intend to develop a model that will be able to predict the outflow for the next 50 years. As a sensible and responsible engineer, you have to perform calibration and validation of the tool. Once a model of the catchment is developed and up for running, you will now have to run the model for the first 15 years 1980 - 1995 and cross check the flood generated by the model with the real time data recorded at the stream gauge point. If the variance is within the permissible limits, then you have successfully calibrated your model. If the variance is very high, you can now tweak the model until the gap is narrowed down. Now apply the same model for the next 15 years and if now the data is closer to the real time stream gauge data, then you have validated your model. Only after this calibration and validation, you are to use the model to predict floods or perform any kind of analysis. If a civil engineer selects a tool without calibration and validation, I say it is the civil engineer who should be held responsible for any harm.

Coming to the next question - NOT NECESSARILY.

If a computer specialist agree to share accountability, it is for sure a much welcomed and appreciated gesture. But if we don't test or calibrate or validate the programs and use it straight for analysis, I'd rather make the civil engineer accountable then the computer guy. He may have written the code but the onus of knowledge rests with the civil engineer alone.

Related Questions

Navigate

• Browse (All)
• Browse D
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.