Rovers and Robots of Aberystwyth University
The Robotics Club at Aberystwyth University (https://www.aberrobotics.club) is holding their Robotics Week in June 2021 and I am one of a number of people who volunteering their time to make 3D representations of some of their research robots.
Each of the models are designed to be exported as glb format for use in the Mozilla Hubs rooms being built. As such I've not gone crazy on details like wiring and all the nuts & bolts.
The rovers Tryfan and Idris were modelled in Blender using reference photos, but not photogrammetry. I used a technique of lining up the Blender camera with an appropriate photo attached, so that floor details match a plane object placed into the scene.
From there it was a process of matching the perspective and then building up a mesh. Locking to the x and y axis as needed while extruding points.
Since they needed to be exported in glb format, I applied the Decimate modifier where I could to bring down the file size. A really useful feature of Blender is that the materials used are compatible with the glb file format, meaning all the surface settings like colour, metalicity, etc are applied.
More information on the rovers and robots is available on the Robotics Clubs YouTube channel https://www.youtube.com/channel/UCEKShXWaYfk4UPefehRcEWA
Information on the rover Idris here https://www.aber.ac.uk/en/cs/research/ir/robots/idris/
Rotation video of the rover Tryfan
Front angle view of the rover Tryfan
Rear angle view of rover Tryfan
Rear angle view of the rover Idris
Front angle view of the rover Idris
Rotation video of the rover Idris
WiP Idris: Initial wireframe fitting while in camera mode. The hazard lines in the photo were really useful for lining up the camera.
WiP Idris: Solid shading of the wireframe to observe progress.
WiP Idris: Example showing using a reference image in the Blender scene to help infer details that aren't visible in the key reference photo.
WiP Tryfan: Initial wireframe fitting while in camera mode. This was more tricky to start off with, a lot to do with the curves but also because the slight fisheye effect the photo has distorting things.
WiP Tryfan: I get carried away and keep forgetting to take WiP snapshots. Here's the upper body section. I like how it looks vacuum formed.
WiP Tryfan: Front console and floor pan modelling done. I had to request an extra photo to be taken showing this part of the rover. Until I saw the photo I had assumed the front console was perfectly mirrored.
The final iCub robot modelled and placed in an environment map from HDRI Haven.
Close up of the arm area before the fingers were modelled.
The iCub robot isn't able to support itself and requires a stand. Again, this was modelled from photos.
Steady progress. Most of the cylinders have been replaced with mechanical components referenced from photographs.
These are the pieces that were imported from the iCub robotics project. Not my own work.
Meshes from here: https://github.com/robotology/icub-models/tree/master/iCub/meshes
The next step is to fit them together and fill in some of the components.