For the first time, CIE has been trying to do a more “team oriented” development pipeline for a project. To do this we started with a simpler one about exploring a model of a bee. For this project, Bart Masters handled the programming, Erica Krieger did the graphic and UI design as usual on top of some additional modeling and texturing, and I was the primary on modeling, rigging and animation of the model. To save on time, we decided to start from a commercially available bee model.

The model we got had a lot of things we were looking for, in particular the overall body shape was good, the abdomen and thorax was separated into proper sclerites and the legs were about perfect. That said, there was still a lot of issues with the accuracy of other parts and changes that had to be made to accommodate the exploration structure we were looking for. A large abdomen object was beneath the sclerites that had to be removed as we would be exploring the internals of the bee and they don’t actually have anything beneath the sclerites like that. The wings were also both inaccurately veined and the veins were physically modeled onto the wings making them very poly-dense in an unattractive way. I went in and flattened and simplified the wings so they could be replaced with just a texture with real wing vein structure. For the sake of our deconstruction of the model, I also had to go through the thorax, which was properly grooved with the sclerites but still a single mesh and break it down into the individual sclerites. This was so we could include a function in the program to “expand” the bee, separating it’s body parts into anatomical segments spread form the center organs for better understanding of how it all fit together.

The biggest issue with the model we got was the head. While the rest of the bee was more or less anatomically correct, the head was completely wrong. It seemed far more like a scary depiction of a bee’s head than what they actually look like. The antenna had the wrong number of parts, the side eyes weren’t big enough, the top eyes were missing, the front of the face had an attempt at getting the facial features right but was still off, and the mouth was just entirely wrong. I almost think perhaps the head was based on some other insect, like a grasshopper or something.

Original Model Head

There was a lot of fixing for the head that had to be done. Ultimately it was basically completely reworked. In particular, the way the mouth parts functioned became a major point of difficulty. Truth is, bee mouths are very odd. They have many parts that overlap over their glossa, the long appendage that reaches out to suck up nutrients. These pieces all fold back into the head and form a flat looking piece when not in use. It took me a few tries to get a setup that looked accurate and fit into the head correctly.

First iteration of mouth parts

On top of all this, we also wanted a “bisected” view of the model where one could see the organs on one half and the bee on the other. The simplest solution here I decided on was to just cut the majority of the bee in half. The mouthparts basically transition into the internal organs so they were the only piece not split down the middle. After all the work updating the model, I had to go in and UV every part of it and put together a rough idea of the color for the textures. This included putting in the wing textures based on an actual photo of bee wings.

Rough Bee Textures

Next came the rigging. We had a specific goal with this rig setup that I had to facilitate, the “expanding” feature mentioned before. Each piece needed to be able to move independently while still remaining as part of the whole structure. This led to a unique rig setup where each part had a specific “detach joint” that could be moved and labeled for use in our program. This is most evident in the sclerites as it forms many joints coming from the root joint that I chose to place in the thorax. Since some sclerites were split in half by the bisecting of the model, a single joint was placed around their center and both pieces were bound to it which lets the piece still function as a single piece while allowing the bisected option.

Bee Rig

Finally it came to bringing the model into Unity. Erica also provided higher detail textures based on the base I set up that were made in Substance Painter. The bee was also given a full smoothing. This drastically raises polycount, but the bee is the only model in the scene so being very high poly is fine. We wanted to use a fur plugin on the thorax and found one that looked very nice, but unfortunately we couldn’t get it to work in WebGL. Perhaps if we make a different build in a higher power system like VR or standalone PC/Mac, we’ll be able to make the bee properly fuzzy.

Final Bee in Unity

For the final step, I did various animations for the bee based on client requests. In particular, they wanted animations demonstrating how the bee uses the mouth parts to eat and how the bee moves it’s legs around the body to gather the pollen it picked up and pack it onto the back legs. I also threw in a flying animation due to thinking before they had to fly to do that pollen packing. Turns out they don’t. When on the ground, they can also pack the pollen, they just sorta stumble around and try to balance on their abdomen and back legs as they move their front legs to scrape the pollen.

Pollen Packing Animation

This project is being handled very differently from previous developments and I enjoyed being able to focus just on modeling, rigging and animation for once. There were some mishaps in transferring assets at times, but those will be used as examples going forward on what to avoid, allowing us to streamline our group development. Moving forward, we’re hoping we can make more projects like this and learn to work better as a team unit allowing us to develop more projects at a time.