Well, you can go down the route of using a proprietary viscometer.
In my experience, most process viscometers measure dynamic viscosity.
This includes capillary, rotational and vibrational types.
It is usually necessary to include density if you want kinematic viscosity. For example, using a process capillary viscosity analyser. This is because this uses the pressure drop at a constant flow rate to determine viscosity. In the lab a capillary viscometer measures kinematic viscosity because it is the time taken for the fluid to flow through the capillary under gravity.
Most vibrating element viscometers only measure dynamic viscosity, the exceptions are the Emerson 7827 and 7829 viscometers and the Lemis VDC 52 ViscoAnalytic. These also measure the density which enables them to convert the measured dynamic viscosity into kinematic.
Non-Newtonian fluids usually result in differences of opinion even between manufacturers.
The argument is that shear dependent behaviour requires a constant shear viscometer (e.g. rotational). However, while in the lab with a fixed sample, the only shear on the fluid is that created by the rotating cylinder or bob, in the process there is the added shear due to flow. Now if we also have a viscous behaviour where work history has an effect then it gets even more problematic.
There are two types of viscosity measurement:
Behavioural where it is the viscosity at process conditions that is important,
Quality or Analytical where it is the viscosity at reference conditions that is needed.
Behavioural measurements are usually those where the product is sprayed, used for coating etc. In these cases it isn't so much the viscosity that is important but some other factor such as reject rates. For example, when lacquer coating headlights, where the lacquer is then UV cured, the viscosity (behaviour) is controlled by adding solvent. There are two objectives, the first is to ensure that the headlights receive the required coating thickness and secondly to minimise the amount of solvent used. In such applications no one actually needs the actual viscosity, just a repeatable value. That is, a value that can be correlated with satisfactory production of headlights.
Analytical measurements are where you want to use the measurement to control the quality of the fluid. In this case because viscosity varies with both quality and temperature (and shear rate for non-Newtonian fluids, and work history with some others such as automobile paints) it is important to control all the extraneous effects on viscosity - the temperature and shear rate, so that viscosity only varies with quality. An example is fuel oil blending. The blending can take place over a range of temperatures so the viscosity varies with both temperature and the ratio of distillate to residual fuels. So either one has to control the temperature to the reference temperature in the viscometer or apply a temperature correction.
These analytical or quality measurements are the most challenging, especially once away from hydrocarbons or Newtonian fluids. However, in these measurements precision of measurement is important and this usually is best achieved by using a slip stream installation with a constant flow rate pump and this then also regulates the shear. So it doesn't matter if it is a rotational viscometer or a vibrating element viscometer (which will have a fixed apparent shear rate which is fluid dependent). With non-Newtonian fluids you then need to correlate the process measurement with a laboratory measurement.
With a behavioural type measurement if the shear rate changes or the temperature changes or the quality changes, the viscosity will change. But since there is a target viscosity at which the behaviour of the fluid is optimal for spraying or coating, then it doesn't matter what causes the change, the viscometer is used to control a heater or solvent addition in order to bring the behaviour (apparent viscosity) back to the optimal value.
In the case of behavioural measurements or some process control applications, the history of viscosity measurement is such that the end users may innovate their own solutions. If all you want is a repeatable value then you look for something that will give you that. It may be a viscometer but it could be something else.
For example, a Variable area meter is one where the float rises or falls in the tapered tube according to the change in flow rate. But if you run at a constant flowrate, then the float will rise or fall according to the change in viscosity. So one approach might be to set up a small diameter slips stream with a small variable speed PD pump and a VA meter. Then adjust the pump speed so that at the optimum viscosity the float is mid range.
The float will rise if the viscosity increases or fall as it decreases. It is now a matter of correlating this with whatever it is that is of interest e.g. reject rate of the finished product. The advantage here is that you are controlling the shear rate to be repeatable by controlling the flow rat using the pump (he variable speed is to find the optimum flow rate. Then leave it alone.... or pick a pump and try different VA meters).
Viscosity is challenging to measure successfully but most of the difficulty is in controlling or compensating for process variations and being clear about what the objective of measuring viscosity is, analytical or behavioural, and whether it is the viscosity that is important or some consequence of viscosity change due to whatever cause.
If you would liek to give some more detail I would be happy to help further.