Case studies on trajectory prediction

The previous chapter discussed cockpit design and integrating novel modalities of interaction. Pointing and selecting targets in displays in different configurations (head down, head up, head mounted) can be challenging. In this chapter, we will investigate technology to facilitate human–machine interaction for direct manipulation interfaces, which allow for directly pointing and selecting targets in a graphical user interface (GUI). Target prediction or intent recognition is well investigated in the field of human–machine interaction for reducing pointing times while operating GUIs for both physical and situational impairment. Earlier work on target prediction worked on 2D coordinates, but later research for touchscreen systems or automotive dashboards investigated 3D coordinates. Situational or physical impairment introduced new challenges in target prediction as the trajectory turns complex and non-linear. However, many previous works did not investigate how the prediction model can be ported for a new task or situation.