For a given part, the machining operations cannot be necessarily performed in any arbitrary order [157]. Geometric and technological constraints will require that certain operations be performed before or after other operations.

AMPS [153] uses heuristic techniques to determine precedence
constraints among features. A number of rules based on machining
practices have been defined and are used to determine precedence
constraints among pairs of features. This approach allows for *
strict* and * loose* constraints. Strict constraints cannot be
violated, while loose constraints can---but at a detriment to
ensuring good machining practice.
The features in this approach are allowed to have multiple approach
directions and may require conditional precedence constraints.

The Machinist system [158] is capable of handling the precedences that arise because of setup considerations. In this system, precedences are generated by examining the setup interactions among features. If machining of a feature destroys the precondition for clamping during machining of another feature, then these two features interact and a precedence constraint exists.

Because of its closeness to well-known combinatorial optimization problems, optimization of operation sequences has received significant research attention. A number of systems have been developed that take precedence constraints as input and find the optimum operation sequence [75,159]. However, most of these systems do not automatically generate the complete set of precedence constraints.

Gupta * at al.* [25] provide a systematic method of
finding precedence constraints which considers tolerances, feature
accessibility, machining practices and machining time.

Precedence constraints are also important in generating and evaluating alternative assembly sequences. De Fazio and Whitney [31,160] provide some examples of that.

Fri Apr 14 13:59:16 EDT 1995