Specs of the product are essential in in the Design Specification Document to ensure that testing and development can be properly executed. The specs of the product must be properly documented to allow for product development in the future. In addition, any product will have a range of error in development accounting for slight variance in size and shape. This must be accounted for prior, therefore when it is sent out to be tested on, the results can be verified. As for determining the specs, that largely depends on the product/device you are developing. In terms of the acceptable range of variance accounted for on your documentation, it is required that all means of failure and error be accounted for. For example, to submit the specs for testing and production of a device that will be 3D printed on a Mark Forged printer, it could be necessary to account for a dimensional inaccuracy of up to 125 microns.

To start off, the specs that should be specified for a medical device should only be directly related to its functionality, as well as interaction with the user. Functions such as image accuracy and scan speed of a dental scanner, for instance, would be appropriate specs to be specified because they directly define the functionality of the product. In addition, the scanner's weight, size, and dimensions are some of the specs that correlates with its usage by the user. Now, the extent of those specs, primarily the functionality specs, must be carefully determined by thorough testing conducted by the company. Companies can only advertise their medical device with the specs that the device can safely and effectively reach. Therefore, determining the true potential of a medical device is crucial and necessary for determining the extent of its specs.

To me, this is probably one of the hardest part of the design process. There are many problems associated with creating tolerances for various aspects of the product. On the one hand if you make the tolerance too tight, it may be difficult for manufacturability or become extremely expensive when ordering from your vendors. When the tolerance is too loose, you get parts which don't assemble right or features which do not work. For example, if you have a hole which is on the lower control limit and a pin which has a diameter on the upper control limit. This combination may make it so the pin does not end up fitting into the hole and there is a failure. This is how tolerance stacking works and can be a difficult calculation sometimes. Sometimes, it all just comes down to trial and error and running endless combinations to see which dimensions for certain features will result in a successful product.

Specs is determined during the test of the product accordingly to that the specs is also upgraded. As the product is fabricated there will be chances of some specs getting destroy in the manufacturing site. So that the reason product goes into multiple tests which comes out as final and better product.

I mostly agree with what what said above. I believe there is a hierarchy of importance. First, the device needs to be within tolerance of it's designed use. You can't have an implant that is too big or small from it's designed size. Next, you need to have the correct tolerances in the parts. If the pieces won't fit together properly, it's not a working product. Lastly, the tolerances allow for machining differences. If you required exact changes, you could make a few in spec, but soon after due to natural changes/differences in the machinery you would fall out of spec very quickly.

From my experience working on projects, deciding on detailed specifications may seem to be the most daunting task, but becomes clear-cut the more you are familiar with the process. The most important factor when specifying ranges of specs is the limiting conditions involved in the design input. I work on designs for a Solar Car team, and when we decide on design specifications, the ranges for design inputs are either based on earlier inputs that we set as a condition to meet or design regulations. For example, FSGP regulations requires that the vehicle must have two inches of ground clearance, so this sets the lower range for many design specs that will go into the suspension systems of the solar car. Sometimes it is based on the objective of what we want to achieve with a next generation solar car. Say, we want straight-line efficiency to take precedence over cornering efficiency in the next generation solar car, then the tolerance for specs that directly affect the rolling resistance (leaning of a car while cornering) can have a wider range without significant concern. Another factor that helps decide on a range is simply following standard equations that are used in specific designs that can help allude to desirable ranges of a design input. There are several design limitations that are already in place that helps deciding on detail specs a lot easier.

In my experience, a lot of specifications have a tolerance enough where the manufacturer is able to make the product well and it doesn't compromise the integrity or functionality of the product. In certain cases, there have been instances in which the manufacturer physically cannot make the range due to their equipment. In these instances, the engineers between both companies would meet and discuss what is possible and what is not. If the conclusion from this meeting is that the manufacturer cannot make this product, then the company needs to outsource to another company that can make the product as per the specification. I've seen that in most cases that the engineers between both companies can come together and make a viable plan to adjust the specification. This issue is rare as the engineers of one company keep in mind to keep the tolerances to a level where the manufacturer can make the product without tedious issues that can hinder production.