Since maintenance instructions are authored by people for aircraft maintainers, reducing errors and instruction update costs are desirable Air Force goals.  Automated Technical Order Validation (ATOV) seeks to use human model models to validate that instructed tasks can in fact be performed under typical maintenance conditions. There now exist several human form modeling and animation systems that are typically used interactively to test human factors and visualize maintenance tasks.  In a series of projects, we have examined the issues surrounding the direct interpretation of maintenance instructions as animated tasks.  A recent Air Force report (AFRL-HE-WP-TR-2000-0088) documented many of the issues involved in maintenance task simulation and validation.  One outcome of these studies is a Parameterized Action Representation (PAR) to support language-based task understanding, analysis, and visual simulation.