Keywords

1 Introduction

Individuals rely on a number of wayfinding devices to assist with navigation. Today these devices most commonly include global positioning systems, mobile phones, maps and compasses [1]. Experienced hikers rely on one or even several of these devices for successful navigation [2]. However, amateur hikers, who do not always own such an array of devices, face challenges when it comes to hiking. This work started out as an exploration to find a navigational solution for amateur hikers. For the purpose of this study, an amateur hiker is defined as one who has some minimum prior experience, but is not an expert with authoritative knowledge of the domain.

In exploring potential solutions, we went through several possibilities. The existing literature in this space and our own exploration into the world of hiking led us to realize that the solution would have to be: (a) non-intrusive (b) GPS based and (c) of a favorable form factor. The rationale for this and the resultant choices made are explained below.

Non-intrusive.

Devices used by hikers help them follow the trail and rely heavily on distracting visual cues. They distract the hiker’s attention away from their surroundings to a map or a screen, which is risky. Auditory interfaces are also not an option as the auditory sphere is a source of helpful feedback for a safe journey. Thus, direct-visual or auditory-cue based devices are not suitable for a mobile user [3]. This finding was central to our discussions and led us to move away from traditional systems of navigation. In [4], the authors found 3D virtual world representation to be quite promising upon comparing different wayfinding visualization techniques on a mobile phone. AR systems take this promise further as they have the capability of overlaying virtual information onto the real world in a natural, non-distracting manner. Thus AR emerged as an interesting space to find a fitting solution. Augmented Reality navigation systems have been discussed before [5], even in the context of AR of terrestrial navigation [6]. We take this discussion further by considering an AR display system that could work in the specific context of hiking.

GPS Based.

While AR showed potential to be the display system, the source of the trail-information was yet to be determined. Traditionally, hikers rely on GPS based data to navigate as these are more reliable and do not suffer from the degree of signal issues associated with cellphones [2]. Thus we decided upon GPS based map data. Reliance on GPS is also confirmed by our study, through the participant interviews and self-reported survey data.

Favorable Form Factor.

It has been found that if technologies for physical activity require the user to carry or wear devices then form factor is a critical consideration [7]. Integrating the technology into the cellphone was one option. However, it is not uncommon for users to temporarily abandon their cellphones with a desire to be unavailable and disconnected [8]. This desire is more commonly seen during outdoor activities like hiking. Further, the cellphone having functions other than navigation, tends to lose battery too soon. Thus, we needed a dedicated device which allowed the user to pursue a disconnect from the world. After considering multiple options we settled on the hiking stick- an existing component of hiking gear. Apart from a form-factor already adopted by the hiking community, it has the affordance of gripping which provides for interesting modes of interaction.

2 Conceptual Design Description

Thus, we present Will ‘o the Wisp: a Hiking Stick with a GPS-based AR navigation system (see Fig. 1). The staff of the hiking stick has a projector that projects navigational cues which overlay on the environment. These navigational cues appear or disappear by the click of a button situated near the top end of the stick. It is designed to be turned on or off by the thumb. This idea was to provide choice by providing on-demand directions and to keep the interaction simple so that it does not interfere with the primary task of hiking. Users would be able to upload a GPS route to the stick beforehand.

Fig. 1.
figure 1

Conceptual diagram

Full size image

In this conceptual design, directions are reduced to simple visual instructions. We expect these clear cues to reduce the mental load required to self-orient and navigate. They overlay on the environment and so, are non-obtrusive. Thus we hypothesize that this design would be convenient to handle while on a hike, has an interaction that is easy to use and learn, and is unobtrusive. We highlight our evaluation goals in the following section.

3 Evaluation

Participants.

Fifteen (15) individuals (6 women and 9 men, ages 18–39) participated in the evaluation, including four (4) subject matter experts and eleven (11) qualified participants. Inclusion criteria required participants to be fluent in English, over 18 years old, able to see (eye glasses or contact lenses permitted), capable of providing written consent, and comfortable walking up to ten feet. Experts met the same inclusion criteria, as well as being Human-Computer Interaction, Human-Centered Computing, and Engineering Psychology graduate students, experienced/advanced hikers and having previously conducted heuristic evaluations. Participants were recruited from the Georgia Tech student and faculty populations, as well as the greater Atlanta area. Evaluations were conducted over a two week period and no compensation was provided.

Prototype.

The prototype was constructed using the following components: (1) lightweight plastic mop handle, (2) clear, laser-cut acrylic arrow, electrical tape, (3) small paper coffee cup, (4) limit switch, (5) 10-foot USB cable, (6) 180° rotating servo motor, white foam core, (7) Arduino circuit, a strip of approximately fifteen neopixel lights. For testing, the fixed Arduino circuit was connected via the USB cable to a MacBook Air running Arduino software.

A 5” foam core platform supported the electronic components. Both the mop shaft and platform were covered with black electrical tape to create a uniform appearance. The shaft passed through a hole in the platform 1” from the edge. A coffee cup covered with black electrical tape covered the electrical components.

The Arduino circuit board was wired and connected to the servo motor. The board was fastened vertically against the mop handle on top of the platform. The servo motor was positioned approximately \( {\raise0.7ex\hbox{$2$} \!\mathord{\left/ {\vphantom {2 3}}\right.\kern-0pt} \!\lower0.7ex\hbox{$3$}} \) of the way down the platform away from the mop handle, with the acrylic arrow attached to its arm. Neopixels were secured with clear tape under the arrow. The switch was positioned on top of the handle, above an existing grip. Wires connected the switch to the Arduino circuit to control the neopixel lights. A 10-foot USB cable attached to the circuit board connected the prototype to a laptop, allowing researchers to manipulate the orientation of the servo motor to move the arrow (Fig. 2).

Fig. 2.
figure 2

The directional components of the hiking stick

Full size image

Objectives.

In addition to task-based usability testing, the evaluation focused on two research areas: (1) Did users trust the novel interaction? What was necessary for them to trust it? (2) How did the efficiency of the interaction compare to existing navigational tools?

Procedure.

Evaluation sessions included four activities. Each session was conducted with two researchers present.

Activity 1 was a semi-structured interview relating to existing navigational tools. The goal was to record baseline familiarity and dependence on navigational technology. Participants were presented with images of four common navigation tools: a compass, a map, a GPS device, and a smartphone, and asked to identify each item and explain which devices they would take with them on a hike where cellphone service was not available.

Activity 2 was a fifteen question paper survey to gather demographic data and information about their hiking experiences, including expectations and equipment. Questions included Likert scales, open response and multiple choice options.

Activity 3 included two tasks: Self-Navigation and Directional Efficiency. The Self-Navigation task was a think-aloud task. Participants were instructed to follow navigational cues provided by the prototype, and discuss their thought process while using the device. Navigational cues were manipulated by a researcher from a distance of 8–10 feet. The Direction Efficiency Task measured the efficiency of identifying in which direction to proceed based on navigational cues versus a traditional map. Efficiency was defined as task completion time. Participants were presented with a map or the prototype in a randomized order. Time was recorded from the start of the task to the point they indicated which direction to go by either pointing or walking in the appropriate direction. (See [10] for similar comparison.)

Activity 4 was a semi-structured interview about navigational devices (similar to Activity 2) assuming the participant now had access to a fully functional hiking stick. The sessions concluded with a debrief discussion.

Pilot Testing and Heuristic Evaluations.

A pilot session was conducted with a qualified participant to test and adjust the protocol and prototype. Iterations to the prototype included modifying the platform to be adjustable to accommodate participants of different heights, changing the arrow from blue to clear, and modifying the switch to react to a single press (rather than requiring the participant to hold the switch to keep the lights on). In addition, four (4) subject matter experts completed the Self-Navigation and Directional Efficiency tasks, and focused on five of Nielsen’s 10 Usability Heuristics [9]: visibility of system status, user control and freedom, error prevention, aesthetic and minimalist design, and recognition rather than recall. Suggestions included smoother interactions when the arrow changed positions and additional features (not developed for testing purposes).

4 Results

Activity 1: Interview.

Cellphone maps were the most common navigational tool used while hiking. Participants also reported relying on intuition, gut feeling, and/or other hikers for navigational assistance. Half of the participants reported using paper maps and compasses. Only one participant reported relying on trail markers.

Activity 2: Survey.

Overall, participants reported that technology in general was very helpful when lost (4.5/5). 66 % of participants reported they currently used a hiking stick. 71 % of participants indicated they would be amenable with a navigational help system being embedded in their hiking stick. 42 % reported previous issues with losing service while hiking (Fig. 3).

Fig. 3.
figure 3

Task performance times

Full size image

Activity 3: Directional Efficiency Task.

The Directional Efficiency Task required participants to determine the direction of travel indicated by both the hiking stick and a map. In all sessions, the response times for the hiking stick were faster than the paper map.

Activity 4: Secondary Interviews.

Participants were asked what they would take with them for navigation before and after using testing the prototype. Responses differed very little. Most participants reported that they would carry the same navigational tools in addition to the hiking stick (Fig. 4).

Fig. 4.
figure 4

Navigational tool preferences (before and after testing)

Full size image

5 Discussion

Objective 1: Trust.

Trust in the device, or lack thereof, was closely tied to a bigger picture view. Participants felt they were blindly following the device without a clearer understanding of their position relative to start and end points. Participants further expressed concern about electrical failures, including battery issues. Participants indicated that they could not trust such a new product without using it more consistently. These concerns led participants to confirm that they would carry additional navigational devices in addition to a functional hiking stick. Due to the high stakes nature of navigation, a low-tech solution was always preferred as a back-up, in the event of technical failures. More information would be necessary for participants to fully trust the device and rely on it for navigation. Information may include confirmation their destination was reached, warning when traveling in the wrong direction, and/or more fluid directional cues.

Objective 2: Efficiency.

In all trials, projected navigation cues proved more efficient than a traditional paper map in assisting participants to identify the direction of their path. Qualitative data showed that the interaction mapped well to participants’ mental models, making it simple to follow directions. Participants expressed that it might “take the adventure out of hiking” while acknowledge the device could be turned off.

6 Conclusion

This conceptual design demonstrated that projected cues may be a valuable form of navigation aid. Integrating this concept into existing navigation technologies may provide additional support for amateur hikers, due to the low learning curve. Additionally, projected navigation cues may be valuable for other scenarios and populations, including those where navigation is a cognitive burden. Potential populations include older adults with cognitive impairments and transportation operators.