Wednesday, 27 July 2016

On the Dilemma of Trust

               Confidence level is a measure of reliability of the system. The higher the confidence level, the more the reliability of the system. In real cases, 100 % reliability is never achieved. This demands the system to be tested, with an appropriate level of confidence under a predefined margin of error.

                         One type of testing employed in engineering sector is the in-service inspection. It is typically a test or series of tests performed on a frequency established by the manufacturer based on prior experience with the system/equipment/component or upon engineering analysis of the probable failure rate for the equipment.

            
                        A good testing program will maintain a record of test results and maintenance actions taken. These data will be evaluated for trends and serve as the basis for decisions on appropriate testing frequency, need to replace or upgrade equipment and performance improvement opportunities. This type of testing allows the system to be working as usual while the test is being performed.

               What is Trust ?

                     When a system behaves within the predicted limits of its defined performance, it is said to be trustworthy or reliable. It doesn't mean the system is 100% reliable or it will continue to work as it had before. When the behaviour of the system is not according to the expected characteristic curves, the user will start to panic. But here is the question. Is the system faulty according to a non biased, inertial frame of reference? In other words, aren't we to blame for not able to correctly understand the system ?

                This is one point of view that can be challenged. First of all, the system is faulty if it doesn't serve our purpose. And this demands we clearly define our intentions about the system. On the other hand, if we adopted the system or if the system is subjected to external forces, its variation from our predicted behaviour can be attributed to our lack of understanding or errors in expectations of the system.


          
               The solution is periodic testing and analysis of the system. The variations in the external parameters must be taken into account and testing must be performed of known and unknown signals. The feedback of the system must be analyzed and new interface models must be made. This will increase our understanding of the system and thus reliability itself.



          

               Applying this to social dynamics, we can see that any relationship that is build on trust is inevitably subjected to doubt. Doubt is not a bad thing. Trust without doubt is akin to apathy. And saying that you trust someone without trying to understand them is in itself, a form of lie. A cruelty such as this should not be committed to the person on the other end.

                 But is trusting someone completely isn't the best thing to do? The answer is, No. Complete trust is not a real thing in the world of organic life. When pushed to the extreme, all living beings will re-evaluate their priorities and choose its own survival. This is why keeping a constant vigilance is absolutely necessary of all mortal creatures.