With recent technological advancements, robots are increasingly deployed alongside humans in dynamic environments. Construction sites are one of these environments, where human-robot collaboration (HRC) can enable semi-automated assembly that uses robots for repetitive tasks and human expertise for more challenging ones. Previous research has not sufficiently explored the use of visual force feedback for assembly tasks under the difficult visibility conditions that are present in construction tasks. This thesis explores how augmented reality can be used to facilitate robotic teleoperation in construction environments through visual cues based on force and torque data. Besides traditional visual force feedback, data from a force/torque sensor is used to estimate and visualize the contact location between a manipulated object and other components. Different designs were implemented as a prototype, and evaluated through an exploratory user study. The study explores the effects of different force direction cues and the presence of contact location cues on assembly task execution. Eight out of eleven users preferred the presence of contact location cues, though these preferences did not result in significant differences in task performance. Based on the current results, operators should be provided with different options to choose from, as the contact visualization may not necessarily be beneficial to everyone, but has the potential to make assembly tasks less demanding for most users.