Personal reflection - Mircea-Ionut Dobre - 293117

 Personal reflection

Mircea-Ionut Dobre

293117

First project - Vuforia

For our first XR project, we had to make an application using Vuforia for a marker based augmented reality experience. This did not go well to say the least. 

We had constant problems with vuforia. At first it seemed quite easy to implement as we were mostly already accustomed to unity from the previous semester but we ran into problems especially when we started building the app going thru multiple versions of Unity and Vuforia before we got one to run. The light in the room was also a problem when the app was trying to get the marker. We also ran into some problems while getting the gyroscope scripts to work with our android test phone. We only had 1 android phone that was not company-locked (companies won't let you install external APKs on your phone if you have a work profile) so we had to use a Chinese phone.

After getting past these problems the compile times on a pretty potent gaming laptop were abysmally high. 

During our presentation, we had a problem connecting Chinese phones to the projector so we could present. We eventually installed a Chinese casting software that came with some viruses so we could present our app as we did not film the video yet at the time

We all worked on the project together in pairs on one PC which is the reason the code contribution can not be seen on GitHub. My main contribution to this project is choosing the topic, asset searching and incorporation, and getting Vuforia to work.

The second project - The trash app

For our second project, we had to make a markerless AR educational app using ARcore. 

We were highly relieved that we did not have to use Vuforia again as we agreed it was extremely bad for beginners like us at least if not for everybody. At first, we wanted to make a shooter game but we had to make it educational so we settled on a recycling game. ARcore was much more friendly to us than Vuforia. We settled to build it on a MacBook as the gaming laptop we used before was disintegrating and we found out the build times on an M2 MacBook are way faster building in less than 7 minutes on average. It was still quite slow as we were constantly changing things but still faster than the chunky gaming laptop. We first tried a Google Pixel but the tracking was quite bad. Eventually, we tried with an iPhone which resulted in way better results as the camera array on the iPhone was way better than the one on the Pixel, especially in dark-lit VIA study rooms.

The main issues we had with this project were settings up ARcore and trying our other phones to get the plane recognition to work better in dark rooms but we had a way better experience than trying to get Vuforia to work.

After we got the app working, the camera had trouble accurately recognizing the plane. We solved this problem by setting the detection to horizontal only, which improved the tracking and allowed us to properly place objects.

My main contribution to this part was topic selection, asset searching&integration, setting up the app, and getting the plane recognition to work.

The third project - Knife throwing (My favorite project)

For our third project, we had to make a VR experience. We were excited about this part as we wanted to play around with VR headsets and we knew that we could find way more resources for this part compared to the others online and take some of the trial and error out of the process.

This was the biggest project overall as we spend a lot of time fine-tuning it as we got to really like the end result wanting to make it nicer and nicer.

For this project, we make a simple at-first-sight knife-throwing game. We also later added axe throwing into the mix to make it more diverse. 

The main issues we ran into while making this app were making the weapon go to the right position when picked up and also the object not sticking to targets after some time or bouncing. For the first one, we went the trial and error way until we found the X and Y mathematics we had to teleport the weapon to compare to the position of the hand while it is picked up to make it look like it was picked up nicely. 

For the second problem, we found out that the area of the weapon that was supposed to stick to the target was too small. 

Also, in the end, we added moving targets which gave us a lot of headaches trying to sync them up as sometimes they would get triggered by an unwanted event.

In conclusion, we really loved this project and I considered it my favorite actually playing it after finishing the game as we got the throwing physics pretty close to the real thing as I went axe throwing in the past I have a point of reference.

My main contribution to this project was topic selection, asset searching&integration, weapon sticking scripts, arena design, the main tester constantly giving feedback to the one coding on the PC, and the teleportation of the weapon to the hand.

Forth project - Final project - Hogwarts

For our final project, we could do anything we want so we chose to go with VR again as we had the best experience with it. After going thru the asset store and GitHub we have found a free repo that had a quite good Unity of Hogwarts out of the Harry Potter series.

We came up with the idea to implement the flying game from the series Quidditch(flying on the broom with the magic winged ball) but we eventually settled on making a flying broom in that part. 

Our other idea was to fill the castle with mini-games like our Knife throwing but we eventually settled on making a Dementor(demon) mini-game where you can shoot a magic wand at them and kill them. 

We had issues implementing the raycast of the magic wand to kill the mobs and also making the flying not go thru objects. We eventually added sound effects when you would shoot the want but they got annoying pretty fast, not using them in the final demo. 

My main contribution to this project was topic selection, asset searching, flight implementation, after the flying we flew thru the entire map examining the model we had.

All in all, I believe the XR course was a fun experience once we got to the VR projects. Maybe some more guidance with Vuforia and ARcore would have been welcomed as we ran into serious problems, especially with Vuforia. I will probably use the skill learned in this course in my job as we have plans for a VR application so it could come in handy.