I love reading about all the 3D printing improvements made in the past couple of years. It really is lowering the bar to participation in this Maker Revolution. Close cousin in the 3D printing family are the DIY CNC machines, new laser sintering projects and more.
However, I’m not sure these really help nudge us toward building large-scale robotic projects. Okay, I agree, some parts will be manufactured using one of these methods, but there are still some real knowledge gaps that need to be filled as well as some industrial engineering and machining challenges. I won’t go on about welding and steel cutting, instead let’s talk about the finer details of locomotion.
It’s inevitable that the Revolution will get to these Heavier Duty issues as it grows, but it’s worth talking about sooner than later to get the cards on the table. This is partly why I started this blog, to help identify the key components needed and, hopefully, see how to work through and tackle them.
There are two areas in particular that haunt any large robot production. They could be summed up as Power and Agility. A ‘mech that lacks the power to walk, doesn’t really fit the bill. Similarly, one that cannot walk elegantly or smoothly is apt to fail. To best deliver on these, the ideal solution appear to be hydraulics and gyroscopic stabilisation, respectively.
Hydraulics are the lifeblood of any heavy-duty industrial machine. While electric drives are becoming more prevalent, I think there are issues still to be ironed out before they can be considered (battery needs, range, power, speed). So I naturally turn to hydraulics because you can drop a nice sized power plant (engine plus pump) onto a platform and start using the power to at least carry itself. The Law of Diminishing Returns still factors in but not as quickly as hulking packs of batteries!
Consider projects like the Open Source Ecology Power Cube – with its goal of being a portal power plant to agricultural machines fits well with our paradigm here.
Need more power? Drop in another cube and connect the hoses. Of course anyone that’s worked with hydraulics knows it’s not always that neat and tidy, but who said giant walking robots were supposed to born in a pristine lab? A machine shop or auto repair hoist is where some of the more real-world work gets done these days.
I like the power cube because I can actually picture dropping one of these into a small ‘mech and getting some serious hydraulic action. A larger diesel power plant is likely needed but how do we know how big it must be? Sizing up hydraulic requirements can be intimidating especially when your “machine” isn’t built yet. Naturally, as you design the machine you’ll want to know how much space to allocate to the power plant. Hmm, chicken or egg here?
Ideally, all this planning would be modelled in a computer simulation. I’ve started to do some simulating in Autodesk Inventor but needless to say it isn’t particularly designed for walking machinery. I’m sure someone more adept at Inventory could figure it out though.
Which leads quickly to the second key component for me, stabilisation, because I need the simulator to model gyroscopic forces/angular momentum to see how it reacts to various scenarios and how much hydraulic power will be needed.
Hydraulics Starter Kit
We really need to develop small “starter kits” for hydraulics. Larger than, say, Lego Pneumatics, yet small enough that a kid can experiment with them in an apartment or school lab without throwing their back out or blowing a 3,000 PSI hole through them. Naturally, cheaper the better.
After some hunting for “miniature” hydraulics, you’ll find there are some being used in remote control models. They are pricey though, maybe pricier than the real thing!
I sense a Kickstarter project coming to put together such a kit. Hit me up in Twitter @minterstellar to discuss…
Part II – Gyroscopic Stabilisation
To be continued…