How to practically and objectively evaluate the Skill of the Operators?
It will be nice if somebody can share Procedure and formats for Skill Evaluation, training need identification and training plan for the Operators in a more practical way.

We don't have all of our procedures and forms in place yet, but we're adopting a system similar to the one we used in the Air Force many years ago.

We rely on a lot of OJT type training for the operators on a particular machine or process type. We made up a training card for each process/machine type. On the front of the card you record basic employee information and list the any compulsory annual training requirements for the individual (fork truck, preliminary review, safety, etc.). These would be dependent upon the general type of employee.
On the back of the card we are listing specific tasks related to the machine/process that the operator will be doing. The card is given to the operator and the operator is paired up with a more skilled operator from that machine. The traineee must complete each of the specific tasks a number of times and have them signed off by the trainer each time.
We figure that repetition is a good teacher. Also, by listing the specific functions of the machine/process typically required different product to be processed through that sometimes have their own unique things about them.
After having signed off on all of the functions by the trainer, the trainee would submit the card to the supervisor/foreman for final sign off to indicate both training and competence. Also, the supervisor/foreman was responsible to evaluate progress periodically to make sue that the trainee was exposed to multiple products with opportunities to employ each skill.
So far it seems to be working out okay for new employees. We have a lot of older more experienced operators. The form is revisable when new or overlooked skills are otr weren't listed. eventually the card becomes a permanent record in the employees personnel file.

Good Luck

Skill-ranking of beginners is somewhat difficult, but in our experience it's particularly difficult to rank high-end skills...for instance, an expertise-ranking system for production processes for which affordable objective measurement methods cannot keep up with cost-necessary production rates, and production is only practical via high-speed judgment of subtle process variations.

We have a number of such processes.

One of course can maintain production records based on off-line testing of sampled parts, demonstrating that somehow expert operator A is more effective than average operator B, but that approach provides rather indirect proof. Plus, if the product for which the stats were kept is not fully identical to the subject product, or if something else conceivably could have changed in the meantime or been responsible for the demonstrated performance, questions can exist.