Open Archive Toulouse Archive Ouverte OATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible This is an author’s version published in: https://oatao.univ-toulouse.fr/24688 To cite this version: Cockburn, Andy and Gutwin, Carl and Palanque, Philippe and Deleris, Yannick and Trask, Catherine and Coveney, Ashley and Yung, Marcus and Maclean, Karon Turbulent Touch: Touchscreen Input for Cockpit Flight Displays. (2017) In: International Conference for Human-Computer Interaction (CHI 2017), 6 May 2017 - 9 May 2017 (Denver, Colorado, United States). Any correspondence concerning this service should be sent to the repository administrator: [email protected] Turbulent Touch: Touchscreen Input for Cockpit Displays Andy Cockburn1, Carl Gutwin2, Philippe Palanque3, Yannick Deleris4, Catherine Trask2, Ashley Coveney2, Marcus Yung2, Karon MacLean5 1 University of Canterbury, Christchurch, New Zealand, [email protected] 2 University of Saskatchewan, Saskatoon, Canada, [email protected] 3 Université Paul Sabatier, Toulouse 3, France, [email protected] 4 Airbus Operations, Toulouse, France, [email protected] 5 University of British Columbia, Vancouver, Canada, [email protected] ABSTRACT dials, keypads and wheels. While computer-based flight Touchscreen input in oc mmercial aircraft cockpits offers instrument displays are becoming increasingly prevalent potential advantages, including ease of use, modifiability, (i.e., the ‘glass cockpit’), these displays are almost and reduced weight. However, tolerance to turbulence is a exclusively used for data output, and user input to displayed challenge for their deployment. To better understand the objects is dependent on a separate indirect device. When a impact of turbulencen o cockpit input methods we conducted cursor is incorporated in the display, a trackball is typically a comparative study of user performance with three input used for item selection. methods – touch, trackball (as currently used in commercial aircraft), and a touchscreen stencil overlay designed to assist The reliance on physical controls is influenced by several finger stabilization. These input methods were compared factors, including the historical development of cockpit across a variety of interactive tasks and at three levels of environments, pilot expectations, and requirements for simulated turbulence (none, low, and high). Results showed safety, redundancy and adherence to standards imposed by that performance degrades and subjective workload regulatory authorities. For example, Boeing and Airbus increases as vibration increases. Touch-based interaction conform to the ARINC 661 standard [1] that stipulates was faster than the trackballn whe precision requirements behavioral requirements for GUI components in a Cockpit were low (at all vibrations), but it was slower and less Display System (CDS). The standard permits only a limited accurate for more precise pointing, particularly at high set of widgets, and it is slow to evolve – the explicit account vibrations. The stencil did not improve touch selection times, for cockpit touch input first appeared in the 2016 update [2]. although it di d reduce errors on small targets at high Touch interaction in commercial cockpits offers potential vibrations, but only when finger lift-off errors had been advantages to pilots, airlines, and aircraft manufacturers. eliminated by a timeout. Our work provides new information Pilots may gain from familiar, expressive, and direct means on the types of tasks affected by turbulence and the input for interaction in certain tasks. Airlines and manufacturers mechanisms that perform best under different levels of could gain from reduced hardware installation complexities. vibration. Replacing hard-wired physical controls with touchscreens Author Keywords could therefore ease development, facilitate upgrades, reduce Touch interaction; turbulence; aviation. weight, and improve pilot interaction. ACM Classification Keywords However, turbulence is a challenge for cockpit touchscreen H.5.m. Information interfaces and presentation (e.g., HCI): deployment. There are risks that the pilot’s ability to interact Miscellaneous. with the touchscreen may be substantially impaired or eliminated during periods of heavy cockpit vibration. INTRODUCTION Previous studies have shown that vibration can be a factor in Commercial aircraft cockpits are replete with physical touch input, but there are few studies that look at different controls, including many forms of switches, knobs, levers, types of interactive touch tasks or that consider how the problem of turbulence might be reduced. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are We therefore examined users’ ability to interact with various not made or distributed for profit or commercial advantage and that copies types of interactive objects at different vibration levels. We bear this notice and the full citation on the first page. Copyrights for components of this work owned yb others than ACM must be honored. used a motion platform to expose participants and a large Abstracting with credit is permitted. To copy otherwise, or republish, to touchscreen to three levels of simulated turbulence – none, post on servers or to redistribute to lists, requires prior specific permission low, and high. All tasks were completed using three different and/or a fee. Request permissions from [email protected]. means for input – a trackball (used in many commercial DOI: http://dx.doi.org/10.1145/3025453.3025584 aircraft), a 21.5 inch touchscreen, and the same touchscreen augmented with a guiding stencil overlay. The stencil overlay consisted of a 3mm-thick transparent Lexan sheet predominantly rely on mechanical switches, dials, and levers that entirely covered the touchscreen (see Figure 1). Holes with wiring redundancy (in case of failure on one channel). were cut through the sheet to permit interaction with The reliance on hardware stems from the aviation regulatory underlying widgets. The sheet’s 3mm thickness eliminated environment, including conformance with standards. capacitive sensing of the finger outside the cut-out regions. While physical controls retain many advantages for certain The stencil overlay provided two potential benefits for flight-critical tasks, there is substantial industry interest in turbulent touchscreen interaction. First, it might offer moving functions to cockpit display systems. Examples stabilization benefits because users can place their fingers or include the Lockheed Martin F-35 Lightning II stealth hand-edge on the stencil, without contact registration, and fighter, which has a large touchscreen display [9], Garmin subsequently move their finger to the target (by sliding over International’s patent for dual touch/cursor control of a CDS the stencil and ‘popping’ into a hole). Second, once within a [22], Rockwell Collin’s cockpit touchscreens [13] for light hole, users can further stabilize movement by pushing jets and turboprops, and the Thales Group “Avionics 2020” against the stencil’s edge. vision for the cockpit of the future [42]. See Kaminani [20] and Hamon [12] for reviews. In empirically examining touchscreen interaction in the cockpit, Stanton et al. [33] analyzed the effectiveness of four different input devices (trackball, rotary controller, touch pad and touchscreen) for menu navigation in non-turbulent environments. They concluded that different devices have different strengths, and that the touchscreen had the highest number of ‘best’ scores. Barbé et al. [4] and Huseyin et al. [17] examined ergonomic aspects of touchscreen positioning in airplanes and helicopters. Wang et al. [37] recently examined the influence of the size and shape (square versus rectangular) of touchscreen targets. Targeting time results Figure 1: Experimental set up. Participants sat on a motion were consistent with Fitts’ law, errors were higher with platform and made selections using a touchscreen or trackball. rectangular targets, and participants preferred larger targets A transparent stencil (shown) was overlaid on the touchscreen. to small ones. In another cockpit-directed targeting study, Results include the following findings: touch interactions Lewis et al. [24] compared left-handed targeting were faster and preferred at low vibrations; the stencil performance when using a touchscreen and trackpad (left- assisted touch selection of small targets at high vibrations; handed performance was studied based on the assumption the stencil increased accidental lift-off errors, but this can be that the right hand would be on a control stick, which is not accommodated by using a short timeout between touch the case for the captain in a passenger aircraft). The main selections; drag-based selections were highly inaccurate finding was that participants were faster when using the during vibration with touch and trackball; and multi-touch touchscreen than they were with the trackpad. Neither of pan-and-zoom interactions were much faster than the these studies examined the influence of turbulence. equivalent trackball method, even at high vibrations. In addressing issues arising from turbulence, at least two We make three main contributions: we demonstrate that patent applications disclose methods to facilitate the use of different tasks have very different tolerance to turbulence;