Robot Arm

I worked at the urological department of UCI Medical Center as an intern for robotic surgery, and it gave me some ideas on how to improve robotic surgery and how to improve its cost effectiveness. For the multi-million dollar price tag, robotic surgery doesn’t currently do what it was cut out to do.

We have yet to see surgeons operating with near telepathic efficiency on patients thousands of miles away. At best it has only been a slight telepresence of a surgeon supervising operators in a faraway room. The Da Vinci robotic system is operated by a surgeon only ten or twenty feet away connected by wire.

Of course this is limited by the heavy and secured datapath that must be maintained over the course of the surgery to effectively carry out the surgery. But given the high speeds of information propagation these days, why isn’t telesurgery more popular? Surgeons should be able to operate in places too inaccessible or distant from their current location with near full effectiveness (with a nursing staff to match of course). And with some improvements in machine vision, maybe even fully robotic automated surgical procedures in the future.


So here’s my first robot arm. I’ve had the CAD files done for a while, but I had no time to actually fabricate this thing until now during winter break. The 4-axis frame sans gripper looked like this rendered in SolidWorks:

The base didn’t  turn out as expected since acrylic is expensive and I didn’t want to waste any scrap. I substituted it with a smaller piece of acrylic mounted on some cut up 1×2’s. Yes wood, we meet again. The frame itself was laser cut out of one (surprisingly small) scrap of 6mm acrylic. The arm frame I designed isn’t exactly the strongest or most stable of all possible configurations, but adding steel servo supports was too expensive for my college budget. As a result the base rotation joint is unnervingly unsupported. I will probably need to add a shaft support later if this arm is going to make any fast movements.


The four servos for each arm DOF are all Hitec brand servos. The gripper and the end effector servo are both from Sparkfun electronics. I was very unimpressed with the gripper. Nothing fit like the site advertised and the medium servo they said would fit needs to be mounted at a strange angle to even fit mounted on the gripper, which seems to be shamelessly taken from Thingiverse (or maybe it is the other way around, which in case I offer my apologies). Either way, the gripper is made of aluminum with decently strong construction, though it is pretty imprecise in its bearings. I won’t be doing any precision gripping with this particular effector.

I’m currently working on some air muscles too, to try to make a Festo style fin ray effect gripper out of molded Polymorph. We’ll see how that turns out later on.

The arm itself is under strain when the position is not exactly neutral, which may be a problem if the arm needs to hold its position for any amount of time. By the way the servos are humming, it seems like I will have to add some counterweight/spring neutralization of the arm’s dead frame weight. Mounting springs between joints would reduce the required torque output of the servos for holding off-centered positions.


So how does this relate to robotic surgery/telesurgery? I suppose part of me at the end of last summer thought it would be cool to create a cheap/simple system that would require little data transfer and still be able to effectively operate on patients over long distances. This robot arm “might” have that capability, but I am highly doubtful the precision is even close to acceptable.

Future work: I plan to increase the functionality of the servos by modding all of them with the Openservo magnetic encoders. That way I can use velocity profiles etc. to make it all run more smoothly. I’m currently making a simple control system that uses potentiometers on a miniature scale version of the arm to puppet the arm. In the future I will use servos for force feedback. The potentiometers connected to an Arduino and an SSC-32 board I obtained for free from my lab’s clean out giveaway + some XBee’s + a cheap video TX/RX system might allow me to manipulate objects from a distance. That would be one step closer to telesurgery right? Maybe even mount it onto Spirit MK2 and have a nice military IED diffusing robot clone.

At very least, I’ve downloaded some industrial arm controlling software if I ever want to make five more of these things and start my own production line.

OR maybe even make myself one of those robot arm helpers Tony Stark plays around with in Iron Man. Oh the possibilities.