Stabilising Touch Interactions in Cockpits, Aerospace, and Vibrating Environments

Abstract

Incorporating touch screen interaction into cockpit flight systems is increasingly gaining traction given its several potential advantages to design as well as usability to pilots. However, perturbations to the user input are prevalent in such environments due to vibrations, turbulence and high accelerations. This poses particular challenges for interacting with displays in the cockpit, for example, accidental activation during turbulence or high levels of distraction from the primary task of airplane control to accomplish selection tasks. On the other hand, predictive displays have emerged as a solution to minimize the effort as well as cognitive, visual and physical workload associated with using in-vehicle displays under perturbations, induced by road and driving conditions. This technology employs gesture tracking in 3D and potentially eye-gaze as well as other sensory data to substantially facilitate the acquisition (pointing and selection) of an interface component by predicting the item the user intents to select on the display, early in the movements towards the screen. A key aspect is utilising principled Bayesian modelling to incorporate and treat the present perturbation, thus, it is a software-based solution that showed promising results when applied to automotive applications. This paper explores the potential of applying this technology to applications in aerospace and vibrating environments in general and presents design recommendations for such an approach to enhance interactions accuracy as well as safety.

Mitigation of Biodynamic Feedthrough for Touchscreens on the Flight Deck

“…As shown in this paper, using task-dependent parameters -i.e., a reduction in the gain of the BDFT model (G BDFT ) in the step task, resulting in an average BDFT reduction of 12% instead of a factor 2 amplification -can be sufficient to render a mismatched mitigation effective again. Potential approaches that can facilitate real-time adaptation of BDFT model parameters are, for example, explicit online estimators for the BDFT model's parameters (Olivari et al, 2014;Plaetinck et al, 2019) or predictive methods based on motion tracking (Ahmad et al, 2018).…”