Kinematic updates

Improved Exo-leg design:

Improved Exo-Arm Design-1:

Improved Exo-Arm Design 2:

Hiatus

Looks like I am going to have to shelve this project until mid-next year due to overlapping external projects.

By then I should have the Load Cell low-cost replacement issue solved.

PV

Actuators and Sensors

Will wonders never cease?

I have recently found a local supplier of high powered Electric Actuators, controllers and damping modules!! (In the vicinity of 500# models). The price is definitely right as is the form factor.

So with this in order locally, I am back to the Load cell issue...
What I have found to date is a system that reads and sends a signal once the Load cell is at at a specific pressure-poundage. This data is sent via USB to an application running onboard a laptop, which then triggers a solenoid to effect the actuator. This is all well and good, however, at $1200 a pop, it does not fit into the model I had envisioned for this project.

There must be something out there that will fill this bill and not break the bank?

Anyone?

Thank you Tim Horton's

Of all places to find inspiration, leave it to me to find it in a semi-awake state at a Tim Horton's lineup, especially considering that I am a die-hard Java Moose guy :-)

Always trying to find the best structural and economical components, while waiting in line, I was drawn to the aluminium arms that held up their monitors. Perfect symmetry, weight reduction holes... integrated elbows via a 2 bar upper arm section bolted into a 1 bar lower arm section. Simple as pie, aesthetically pleasing and strong as hell.

Material is 6061 T-6 Aluminium
Dimension Choices:

• Main spans: 1/2" x 1.5" x length

Extremity structure settled. Thanks Tim :-B
_______________________________________________________

PS - After re-reading some best practices on Exo-Frames, I believe I have had a Eureka moment in regards to some of the recommendations from the Pacific Rim.
They recommended a 4-bar stability system for the extremity spans. Up until now, I was unable to get a solid answer from anyone as to exactly what this entails. It suddenly dawned on me while designing and re-designing the extremity spans that this means tandem bars up, and tandem bars down (instead of my single bar in the lower position).

To accomplish this, I would need some sort of Delrin-washered Aluminium spacer at the elbow to maintain the proper gap. The upper and lower insertion (body connection) points can act as their own integrated spacers (includes Delrin washers to reduce friction).

Upon further consideration, I think this would also prove to be a superior platform when it comes to mounting the actuators in a stable manner.

Why PAM turns me on :-B

Looks like Electric Hydraulics in Canada for under $1000 an actuator is a tail-chasing adventure. There are a few on EBay from time to time, but matching them and getting them to work together is more of a pain than it is worth.

I have since chosen to actuate using P.A.M. (Pneumatic Artificial Muscles). The beauty of this solution is the ease of construction and the way they mirror their bio-engineered counterparts. I can actually build these inexpensively, however, the difficult part will be the valve system that acts as a 1 – 0 Boolean switching system. As pointed out to me, pneumatic spool valves might be the answer... and also using the exhaust from one to move the other. More to follow...

For instance, we will need a biceps – triceps inverse contraction relationship. Only one can contract at any given time, never both. Using this type of valve system, it may be possible to hold a contraction at any point of a lift phase. How does this figure in? Well, this sort of contraction – relax relationship is how we are able to maintain our standing up straight stance.

This system could be powered by anything from a pair of SCUBA tanks to a small on board air compressor. Ideally I want this to be a variable-pressure power assist system, not unlike a SCUBA diver’s regulator that gives more air depending on demand, or power steering on a car.

As far as the Operating System goes, the leading contenders are:

-Redhat Linux
-Ubuntu Linux
- Windows XP (There, I said it )

Some if the integrated features will be:

-USB System Temperature Sensors
-Load Cell status indicator (Via custom Widget)
-GPS Positioning (Via custom Widget)
-Laser Rangefinder (Will be building this from an instructable)
-System fault indicator (Via custom Widget)
-Secure Wireless connectivity (802.11g WPA-2)
-VOX command system (Currently testing Smart Butler)
-Emergency system shutdown
-etc

This will all be condensed into readouts and widgets on the 800 x 600 pixel display on the (still in development) HUD.

On Actuation and other solutions

"010001110111001001100101011001010111010001101001011011100110011101110011"

Currently, I think the most frustrating part of this adventure is trying to source the components that I need, only to discover they are in Sweden, or elsewhere in the EU, and they either don't ship to North America or the shipping is horrendous.

Regardless of these setbacks, I think I have settled on an Electric-Hydraulic Actuation solution. It is a beautiful balance of high power and step-motor control... now all I have to do is source them here in North America... getting closer every day.

I have also settled on a I have made some interesting discoveries Exoframe-wise as well:

1. There needs to be a compensation mechanism in place to accommodate the shoulder axis upward shift when the arms are raised to the sky - solved


2. I need to reduce the weight on the distal segments of the extremities to minimize energy cost: I recently came to the conclusion that the knee does not actually have to be actuated! I can get away with a Knee Acceleration Damper called a Dashpot which I can set to have a free swing forward for energy conservation, but proportionally to load, slow the flexion on the way down in order to keep the knees from buckling. Also the ankle assembly is going to be passive and incorporate a sort of mini-McPherson strut system to store and release energy. I will also rig the Exo spine to have the ability to slightly rotate to facilitate the Contra-Postal swing of the shoulders/arms while walking (Which again conserves energy). The arms will only be actuated for shoulder raise and arm curl/press contractions. The forearm will incorporate a rigid rotational coupling for forearm suppination. This will have some mechanical resistance inherent as opposed to a freely rotating assembly.
   

More later...

Creating a HUD

I need some method of projecting an image to the inside surface of - say - a visor. I've been looking at several options including microprojectors, but they just don't seem to fit the way I'd want. If this is going to mount inside a helmet, then I can't deal with any bulk. It has to be able to mount either on the forehead or between the eyes above the eyeline and project (with about an 80 degree FOV) on to the visor, while remaining completely unobtrusive to the human vision.

On top of that, I need to control the focus of the image such that the eye can "see" it from about 4-5 cm away. If you stick your hand in front of your face, you can't focus on it. I'm seeing a lot about focusing on infinity in other forums and such, but I haven't been able to try it yet.

Liquid Crystal on Silicon is another option as well. I need to determine what level of projection these provide, as well as a resolution. This will need to be a fairly wide image, but vertically narrow.