AR Plug-in Comparison
- MASTERS PROJECT
​
PROJECT OVERVIEW
Original Title: How can I create an immersive/entertaining AR escape room?
Over the course of this project I want to use previously gained knowledge about AR try to build on it by expanding the concept and idea of how AR can be used in the sense of immersion.
​
Immersion is user subjective as to what it means to them in a gaming context. I am looking at whether putting AR objects to interact with the player in an escape room scenario would create a better and more enjoyable experience for the user/player than without them.
​
I think AR and VR are both fascinating development tools for the future of the games industry, creating a big step forward in the case of immersion and UX. Currently, it seems that AR development for gaming has slowed after the initial success of Pokemon Go; There are exceptions, for example the Microsoft HoloLens, which is probably the most advanced AR piece of technology available to the public consumer.
As far as the immersion in my project goes, although a headset usually provides a deeper UX experience in an escape room scenario, the idea is to use a phone as a scanner to find hidden objects.
​
Testing Scenario
The finished project will be a short escape room scenario which the tester will play through until completion and then will be asked at the end of the process to fill a survey to identify if they found the experience immersive or not. The test will take place in a set room with all the objects that are required within it. Some objects will be able to be scanned and some will be placed to be used to complete the clues found . The user will take the provided phone and then scan over the objects to find the clues to accomplish the escape room.
​
IMAGE RECOGNITION USE
HOW IT IS BEING USED
The most used process in my design will be image targeting/recognition on items which will be found in the experience. The mechanism of image recognition is that as long as an object has a flat surface, it can project images or objects onto that flat surface.
CREATION
All the image targets were created by taking pictures on my mobile phone which were then uploaded to the Vuforia database online. This database has a function which will automatically notify you whether or not a picture being used has enough distinct features to be able to be found in the real world. I am creating all the images being mapped onto the objects in Photoshop; these are either clues in written form or as pictures to be solved.
PROJECT MANAGEMENT
To stay on track I decided to use some different techniques as aids. I am using a Kanban chart, Ghantt chart and to keep notes I use Microsoft One-note
My workflow
Following on from the last section on project management, I thought I would explain my workflow practice; how the idea progressed from concept through to development. All ideas are jotted down in a book and developmentally, become reality from working in the Unity engine. Some of these are shown in the photos below.
SMALL FEATURES I CONSIDERED
C# CODING
This was one of the hardest parts of the design of the game as I am not coding proficient, and have tried to steer clear of it as much as I could by using image targets. However, to create a compelling game I found that I did need to include some gameplay mechanics and so I took the concept I had from last year and used my maze game as a template where I had written small amounts of code.
VUFORIA
I have touched on it briefly but I am using the Vuforia plugin in affiliation with the Unity Engine to create my AR objects and images with the real world. The image tracking for Vuforia is in my opinion the best option for my needs in this project.
POST PROCESSING
Post Processing is the instance of adding filters to the screen of the game to add different effects to the gameplay.
​
I used post processing on the camera in the AR world to add to the immersion to make the player feel like they are holding a police scanner to find hidden clues.
MY WORKFLOW
In continuation from the last section on project management, I thought I would explain my workflow practice; how the idea progressed from concept through to development. All ideas are jotted down in a book and developmentally, become reality from working in the Unity engine. Some of these are shown in the photos below.
Video Demo
SETBACKS (COVID-19)
What can I do to overcome the situation?
So far (end March 2020), Covid-19 has thrown doubt onto the testing phase of the project, where one to one testing is required in a single test room for consistency and viable results. Whilst the development of the project has not been affected as a result of the infection, it is the uncertainty of the next phase which has made it difficult to quantify. The testing requires human contact and as currently this cannot be achieved, with no date as to when it may be, I will need a back up plan. With the nature of the AR escape room I have built, this provides a major obstacle. As the room utilises specific target objects, I am unable to package the game for others to download and use whilst in isolation, even with accepting that each users room may be different and so not providing exactly comparable data. If I change to items which can be emailed, such as the maze, this will not provide the UX I am aiming for and again, will defeat the object of the proposal. If the enforced lockdown continues then I would be left with asking for an extension to complete the testing fully and so allow proven results or to reduce my test group to just myself and those in isolation with me; this would provide no usable data in a test sense as the group is far too small. Having already spent months designing my test, starting again would bring its own issues.
In the current situation , I am still looking at other ways to continue and still provide a comprehensive test, whilst staying somewhat true to the original concept, utilising my development so far. As the testing is the biggest issue, I am considering testing a different question by myself. This would involve a change in concept from looking at whether the addition of AR to an escape room be an effective experience, to instead look at workflows used to create this experience; Posing a question such as which is the best way to do this or even if it is possible for such a creation using Vuforia and Unity.
Update Project Overview
Comparing the Use of the Unity Game Engine Vuforia and ARCore Plug-ins for Developing Augmented Reality Mobile Games
After the onset of Covid19 and the challenges it produced, I was forced to change my idea from an escape room to something which did not require human participation. The reason for this was that even if I could get a small sample size it would not have provided a large enough test and project to provide enough quantitative data to form viable conclusions from.
​
What I decided to execute was a comparison of the two most commonly used AR plugins in the Unity Engine, namely ARCore and Vuforia. This will allow me to continue with the creation of a technological project utilising AR whilst providing self testing and conclusions using qualitative findings.
​
​
One of the driving forces in this choice was that when developing previous AR apps, I was always uncertain as to which AR plugin I should use for the project or idea I was trying to progress;. In choosing to create two different apps using two different software's, I believe that I can identify which of the two apps is best in different scenarios and what work best with certain AR capabilities.
​
​
​
​
Basic Comparison
MY WORKFLOW/TIMELINE UPDATED
I had to create a revised, updated timeline so that it actually represented what I was going to work to in my final stage of the development. With so many things changing in the whole process, this new timeline was crucial in order to ensure I achieved what was necessary within the time-frame left.
I decided on using Photoshop to plan my workflow instead of just writing my ideas down in a book. This, more practical approach, allowed me plan out my ideas by drawing them on the virtual canvas which made it easier for me to envisage them.
Comparison Conclusion
ARCore
The outcome of the project was that, in my opinion, ARCore is far more stable when it comes to world anchors and plane tracking, it also goes as far as creating a simple visual tracker so that the user/developer can see what has already been mapped. ARCore also allows many of the features to be changed easily in the engine, such as the images being tracked and also the ability to only display on item at a time while plane tracking. I found this very helpful when developing the prototypes.
​
ARCore requires a lot of coding knowledge to get anything working in the Unity Engine, however, as Unity as a base itself also requires coding knowledge to work , I am not using this too harshly as a con in judgement of it. I did find the UI of the app to be very confusing and overly complex, there were many drop downs to go through just to find the element that you needed. In my opinion the image tracking in ARCore is not very good at this stage; This may have been because of the image I used, however Vuforia still managed to track it without a problem. Also, the inability to track two images at the same time was a large downside to the whole development.
​
It was a shame that I couldn't get the image tracking game to work (Left video) but I was very happy that at least the plane/anchor tracking game work finally in the end (Right video)
​
ARCore Videos - Recorded on Google Pixel 2
Comparison Conclusion
Vuforia
In Vuforia I found that Image tracking was the best element involved in the use of plugin. Being able to track multiple images at once and and keeping them steady when doing so is a big plus. The world anchor tracking was also stable but the more that was put into the scene, the more lag the game produced and the worse it tracked.
A big positive when using Vuforia is that far less coding was required to make the app work. For me, this is a big plus.
With many of the functions required to use Vuforia on a website and not in Engine, it is very tedious to import the image datasets to work with image tracking. Although it does not take too long, having it in engine would speed up development a great deal. In my opinion, Vuforia can only be used to create very basic AR applications at the moment .
​
My discovery was that Vuforia allowed me to create a game with image targets (Left video) whilst I was unable to do it with plane recognition (Right video).
​
ARCore Videos - Recorded on Google Pixel 2