Robots can greatly alleviate physical demands on construction workers while enhancing both the productivity and safety of construction projects. Leveraging a Building Information Model (BIM) offers a natural and promising approach to drive a robotic construction workflow. However, because of uncertainties inherent on construction sites, such as discrepancies between the designed and as-built workpieces, robots cannot solely rely on the BIM to guide field construction work. Human workers are adept at improvising alternative plans with their creativity and experience and thus can assist robots in overcoming uncertainties and performing construction work successfully. This research introduces an interactive closed-loop digital twin system that integrates a BIM into human-robot collaborative construction workflows. The robot is primarily driven by the BIM, but it adaptively adjusts its plan based on actual site conditions while the human co-worker supervises the process. If necessary, the human co-worker intervenes in the robot's plan by changing the task sequence or target position, requesting a new motion plan, or modifying the construction component(s)/material(s) to help the robot navigate uncertainties. To investigate the physical deployment of the system, a drywall installation case study is conducted with an industrial robotic arm in a laboratory. In addition, a block pick-and-place experiment is carried out to evaluate system performance. Integrating the flexibility of human workers and the autonomy and accuracy afforded by the BIM, the system significantly increases the robustness of construction robots in the performance of field construction work.