@zeezeemoomoo @ill_wil Then don't have a head at all. Put the cameras all over the body instead. They are cheap.

@zeezeemoomoo @ill_wil Just put it in the trunk. Cables don't have the same limitations as nerves, and processing needed to fuse all this sensor data together is small even by today's standards, not to mention the high-tech standards this robot obviously has to meet. As a bonus you get a stabler base for the sensors (less software stabilization needed) and the whole thing becomes easier to armor properly, with inclined surfaces deflecting shaped charges. You could even have a jacket with additional sensors.

@zeezeemoomoo @ill_wil Yes, that works. I think that's because it is close enough to the actual design process that created it, focusing on looks first, and adding functional details as necessary.

@deshipu @zeezeemoomoo lol, listen it's a robot dog. I don't think multiple cameras are easy to compute based on how fast it needs to react. It could use maybe one more camera but I was more or less thinking it uses sonar which would be less energy consuming. Now the cables act more like muscles, they're a type of fibre that contracts based on a small pulse of electricity. Now you can continue the debate lol

@zeezeemoomoo @ill_wil Then don't have knees at all. That should confuse and distract them.

@deshipu mid section is a joint as well as battery. Everything else is just to give it a animal look. Feet are small because I got tired of drawing lollllll

@ill_wil Ah, so the light is a battery charge indicator? Neat. Wouldn't it be great to design it so that it could be actually built and make sense?

@deshipu eh, it makes enough sense. The frame is tiger/dog based. Most of the power is in the front and you can't add a bunch of cameras because the computer would be multitasking too much. The feet could be small so it can stick into rocky terrain. Other than that, I only did this in an afternoon. I wouldn't call this a final design :P

@ill_wil I get that it's just a sketch and so on, but I think it's a little bit backwards to first decide how it looks like, and then retrofit explanations. You can't add sense later, you have to start with it.
I'm a bit triggered by non-functional mech designs, because I did a little bit of drawing, and also built a few walking quadruped robots, and it immediately jumps out at me when there is a glaring problem in the construction, or when there are parts that have no purpose and only serve as decoration.

@deshipu eh, there isn't much decoration in the design. I'd probably have to rebuild the shoulder area and the feet probably need more rework. Otherwise this would be considered a direction, from here I'd probably start figuring out the design further in 3d.

@ill_wil Sorry for complaining, it's really great, much better than the average mech you see out there — that's why I bothered with the remarks. Sorry if it sounds negative, keep up the good work!

@deshipu hey, it's no problem. I don't mind nit picking and what not, I don't consider this work final so it'll help with any future designs I think.

@xenthreaded Scott Robertson is the number one place to go. His perspective book is top notch. Completely worth getting. Other than thaaaat, practice drawing random mechanical shapes, start by making fun little boxes, cylinders, different joint systems, etc. This will help you practice your perspective while also thinking about design. YouTube '70/30 design' and use sketchbook Pro, it has the best perspective tools, makes it very easy to check your work but try to freehand a bit.

@ill_wil

Thank you for replying.

I'm taking notes of this :)

@xenthreaded no problem! Just start small and work your way up. I still struggle with things like spines or feet myself so don't feel like you need to work on giant mechs right away!