The task of repairing or maintaining complex machinery in a remote collaborative environment can be very challenging for the involved parties, both on- and off-site. Augmented Reality can offer a more immersive experience in this situation for both users. Using both sparse and dense Simultaneous Localization and Mapping (SLAM) techniques, two different systems for remote collaboration were developed. Annotations are provided for the on-site user, and the task environment is re-created for the off-site user. These system allows complex remote collaborative tasks to be conducted more effectively. The main contribution of this research was the investigation and application of various technologies and algorithms to build working prototypes to satisfy a set of requirements.