i am not a robotics guy. i was directed here from “practical machinist forum”. my son has recently been released from a hospital where he spent 7mo w/life threatening burns. he is now able walk some but has no use of arms/hands. he has given me the task of coverting his golf cart to foot/leg controls. (all reasearch on existing handicap stuff relates to hand controls only)
i need some suggestions for steering. linear actuators was mentioned but the supplier says no as they are 20% duty cycle.
any ideas?
thanks, cliff
Best wishes to you and your son.
A hydraulic ram steering system, typical on 4x4/off-road vehicles, would be very effective but are moderately complex and heavy. Could be foot-operated with either a lever or pedal.
Where are you located? I would wager that you aren’t too far from a FIRST team, and they would probably be willing to help you directly if you wish.
i’m near orlando, boy & cart are in virginia beach
OK, so do we find you a team in Orlando, a team in Va Beach, or both?
Also, before we start: We are a bunch of practical guys, some engineers, some not, and a LOT of High School kids, here on this forum. You and your son need to be the final deciders if something is safe or not.
OK, let’s understand the problem: He has some use of his legs, but not his arms or hands. That means to me that he can use throttle and brake, but not steer. Is that right?
Going on that assumption, if the right foot is go/stop, the left foot needs to be steering.
Problem #1: How does the left foot control steering. That is, does one use a gas pedal type control (press for right, pull for left) or a side-to-side lever? I would hesitate for a push/pull arrangement, since a big bump could cause an inadvertent turn. But, think about your son’s capabilities and let us know what you think.
Problem #2: OK, we have some kind of input, how to we physically move the steering wheel?
A linear actuator directly on the steering box sounds nice, but as you said most are really not suited for that - but check McMaster.com and Google for other styles that might be workable.
Pneumatics are right out, since they can’t hold a given position (Air compresses)
Hydraulics are fine, but need a pump for pressure and some plumbing. You don’t need a lot of power (= thin cylinder, maybe 1" or so), but the controlling is pure mechanical.
Any of the above require a permanent modification of the cart, and people without a special skill can’t drive it easily.
Another idea is to install a sprocket onto the steering wheel, and use a motor to turn it via chain drive. Make a method of de-energizing the motor and anyone can steer it. Electronic controls are light and cheap. I like this idea the most so far.
For a motor, look for a large servo motor. A servo motor is a special motor that provides feedback for its exact position. Using some basic electronics you can command it to, say, 7.2 degrees right and it’ll go and stay there. The controls can take several forms, so as he improves different control schemes can be considered, not just a lever at his left foot.
Thinking off topic a little, he might even be able to control this with his head - turn right to steer right, and so on. Does he have good control of his neck?
OK, enough speculating for now. Let us know about #1.
Oh yes, we also need to know a little about you: Your experience, skills, machining capabilities, and budget.
I can think of a couple ways to do this.
Foot steering could be accomplished by using a turntable (torsion spring return to center) that you put your foot and twist to turn, and having an absolute encoder or potentiometer give the feedback to a control system that controls the electric motor.
The actual steering could be accomplished through an electric rack and pinion setup - you may be able to get one cheap off the net for existing cars.
Or, conversely, you could make your own rack and pinion setup.
Or, you could run a small motor in a transmission to a winch with wires running to the end of each A-arm. This is probably the simplest as you wouldn’t have to worry about any type of mechanical alignment with the steering system and a rack.
I’d also look at RC cars. They use a servo with a few simple lever-linkages for steering. You could do the same here, and use an electric motor with feedback through a small transmission as a pseudo-servo.
Finally and probably the most simple would be directly turning the steering column with a motor through a small transmission. You could use a lever and a super-shifter transmission so that even has a neutral gear so that you could hand steer it!
To tell you the truth - this is all something that many of the First Robotics team here could pull off with their machining facilities. Maybe you could even hit up someone for one of the old IFI control systems to program it all for you.
Edit - it looks like Don and I had a lot of the same ideas, but I just thought of another one. Instead of a turntable, use a lever that turns, similar to the way the hand on a clock turns. Put a shoe-catch like they have on long-distance bikes that the toe of your shoe goes in. Then he rocks his ankle left to go left and right to go right. This would be easier than a turntable though it would work on the same philosopy. It would be pretty easy to make.
I think the most important part will be the user interface, not the implementation. Take the time to figure out what is most comfortable before moving forward. Assume that anything is possible (at(in?) first).
I’ve seen this done “in real life”. One of our patients who had had both arms amputated drove his car this way. One foot for gas / brakes. One foot for steering with a turn table. Not sure what the linkage was.
http://www.velovision.co.uk/mag/issue11/specialneeds150.pdf
This is actually from a cycling magazine, but page six mentions hands-free steering and provides a web address. A similar interface could be used even if you keep the same steering mechanism, it could potentially be interfaced into the existing steering column with control cables.
Two important points.
With the turntable idea: Imagine if the steering wheel was on the floor. Wouldn’t that be pretty good? Maybe that can be done with a sawzall or something relatively simple like that.
But Eric’s point is critical: What is best for the user? We need to hear more about that before we go much further.
how about 2 gas pedals, one for the left wheels, and one for the right, then you have a tank drive, it might require some major changes to the golf carts drive train
Take away the “might” in that. Is it doable? Yes, but not easily. (Golf carts use steering like cars; modding that would need a second motor and locking the front wheels at minimum, let alone control setup.)
Depending on the user’s ability, and on the golf cart’s steering column/body surrounding the steering column, I would actually go with putting a second steering wheel onto that rod at foot level, if at all possible. It’s simple, it doesn’t affect usability if he wants to be driven instead of do the driving, and it may actually be pretty cheap.
First, Does your son have no hand function at all? My father had a progressive brain disease that left his body frozen. We got him an electric wheel chair the hover 'round. Its joy stick interface is extremely well implemented and could be used if there is any hand function at all. Well engineered product. I have also seen wheel chair control for people with no body function implemented with a bulb for velocity. The harder you bite, the faster you go. Turning was controlled by head movement. You look at where you want to go. There are many commercial solutions on the market that would be safer and a whole lot easier than a golf cart kludge. However, even with insurance you will have to lay out some substantial money.
Tiller grasped between the knees? would leave feet free for accelerator/brake, might be more natural to drive?
I’d spend more time seeing what’s already available, most (but definitely not all) of the good ideas have been thought of…
You could still use rack-and-pinion steering with such a scheme; the input would just be tank drive. You would have to come up with a mechanical way to convert between the two representations, though, unless you consider drive-by-wire solutions.
Converting the tank drive commands to velocity and steering would not be too hard.
Velocity = (L + R) / 2
Turn = (L - R) / 2
Though, the cart won’t react the same way as a First robot.
Mainly because the cart can’t turn in place, only in arcs.
I would keep the steering control type as close to the original as possible (ie a turn-table would be fine because it is still a type of rotary interface that has a one-to-one relation with the cart’s steering, like the original steering wheel)
If it’s a RWD, add a motor. If it’s 4WD or front-wheel drive, it’s a lot harder.
It’s simpler if you can just add a second steering wheel near foot level.
what about an arm or arms directly attached or geared to the steering wheel that allow it to be controlled by moving one’s knee from side to side.
a Y shaped bar that straddles one’s knee is both cheap and effective gearing could be used to amplify inputs, since a direct connection wouldn’t be very agile without having to lift your knee, and then you wouldn’t be able to operate the pedals.
I’m the one who steered Cliff to this site, so let’s do what we can to help out. I’m sending Cliff the contact info for Patty Cook, who can no doubt hook him up with some of the team contact info.
i can deal w/anything mechanical inc fab & machine. i’m just back online from the weekend and am absorbing the posts. i’m going to investigate the tricycle thing. the boy might not like it but sounds intriguing to me (work on lower body while commuting). his idea for a golf cart is because he already has one. thanks for the ideas, cliff
I agree, if there’s a way to get him some aerobic exercise in the deal, it’s good. But we don’t want to tire him out just yet.
Instead of complex electric motors, sensors and and chains and stuff, maybe just a linkage between a foot plate for steering and the original steering shaft. Left foot steers, right foot gas/brake. Simple is usually best.
Let us know how we can help.
well the aerobic thing is out as i found out he has no sweat glands (only on his butt which didn’t get burnt) and has to monitor temp always.
looked @ the cart this weekend. 3turns lock to lock, using a rack/pinion connected to a bell crank. not alot of effort to turn.
my thinking is a linear actuater connected to the bellcrank would be real easy but the duty cycle and the fact that it would not be easily converted back to manual operation is a concern
coming from an automotive background i thought about splitting the mast jacket, adding a sprockett to the strg shaft, sandwiched between 2 collars and a chain drive to a wiper motor or window motor. bushing the driven sprockett and securing with a removable pin or a gear dog. that way it’s easy to change for normal steering.
fab a footpedal w/springs and stops to control range and effort. weld 2 pieces of bar stock forward and aft of center to operate a windlass sw (reversing sw)
other issues:
old style dimmer sw mounted on firewall (w/lighted indicator) for on/off.
makeup for/reverse arm, hanging down w/handle for leg control
lengthen the existing rail on driver side of seat and adding a safety net from rail to roof (driver unable to fasten seat belt, etc) enter from pass side
what do you think, too easy?