Industrial Robot Hacking

[Ross Bochnek]

My understanding is that the main obstacle to playing with the Industrial Robot is the power requirement, which I believe is 3-Phase 220 Volts, correct?  

Can this be supplied by a reasonably portable and existing or obtainable generator that we could run outside during halfway decent weather? 

I would love to get the robot working and bring it to Ingenuity Fest.  I know the power at Ingenuity Fest is pretty unreliable, but perhaps arrangements can be made.

Also, what does the Tesla Orchestra use for portable power?  Could we possibly arrange something with them and run the robot at Ingenuity when the coils aren't running?

What are your thoughts?

[Jonny Hall]

As a member of the Tesla Orchestra -- We don't use portable power.  We make arrangements with the event organizers wherever we will be performing to make sure there will be enough power available.  Or course, we use 100A, so we are usually connecting right at a breaker panel, and work directly with an event's electrician to make sure our requirements are met.  And yes, there has been at least one event we were looking at but in the end didn't participate in due to power limitations.  There was also one event where we were supplied by a large trailer-mounted generator, but again that was supplied by the organizers, not us.

Comparatively speaking, the robot arm takes a quite small amount of power.  As I remember when I hooked it up once, it was along the lines of 20A 120/208 3-phase --  but realize that to get 3-phase you're likely talking a sizeable generator.   I would recommend that you just get a long (couple hundred feet), adequately-sized three-phase cord for the robot arm. Most events are going to have power available -- you don't see all the food vendors, etc. running their own generators; if there's a generator it's a massive one with distribution around the event. So the power is likely there and available, it's just a matter of communicating with the event organizers and getting access to it.

  -Jonny

--
You received this message because you are subscribed to the Google Groups "Makers' Alliance" group.
To unsubscribe from this group and stop receiving emails from it, send an email to makersallianc...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Ross Bochnek]

Thanks Jonny! So you think 3-phase 120v would do it? I've seen some pretty small gennies that could supply that. If 240 is required, maybe we could use 2 circuits. Perhaps at an event, a 20A stage plug would work, perhaps with our own 60A stage plug adapter or twistlock for distro flexibility? Or, are we mostly looking at twistlock; either 4-0 tails or something with a "washing machine" connector? It would be great to have genny to run in the N side lot for testing/programming at LH. Then we could also have a backup for events, while trying to use their shore power or big generator. I also wonder if the fridge-sized robot controller could be brought to an event, or if we could make our own, more portable one.

[Matt Hummel]

If we were to bring it we would need to position near that power source since they normally don't run 3 phase out along the spyder distribution boxes. Also, we would need to build some sort of base to mount it on so it didn't topple over while in use, as well as a safety perimeter. In addition, we would have to figure out the logistics of just getting it there, is heavy enough that an engine hoist and forklift jack will be need.

Not saying it's impossible but there are some definite things we will need to work through to be ready. But with it being in September we still have time to figure it out.

[Ross Bochnek]

Excellent points, Matt!  I concede that making the robot not only turn on, but also movable AND controllable in some interesting way is indeed a LOT to pull off in 6 months.

There already seems to be a lot of interest among out crew to get this puppy working, and I think Ingenuity is exactly the kind of event that needs a hacked industrial robot, and the motivation and possibly resources we'd need to make it happen.

We'd need to attack operation on at least 2 fronts: Power and Programming.  The main question in my mind is; can we power up the bot soon enough to be able to program it to make it all worth it?

Using it would indeed require 2 more fronts: engineering it to not tip over while it does cool stuff, and making it portable.  Plus, it would require transportation, special placement and ample space.  Maybe we could permanently mount it on a trailer.

Within a week of the conclusion of Mini Maker Faire, applications for Ingenuity Fest are due, including a budget.  At the Tuesday meeting on 4/1, I hope we can put the budget together, and I'll help with that and the rest of the application.  I think this project is even worthy of considering a crowdfunding campaign, and some backer reward levels might include being able to control the 'bot.

I can think of lots of cool things to do with a working programmable industrial robot, and so can you.  I'll take a hint from Tesla Orchestra in understanding that we'd probably want to bring such a project to more events that could actually fulfill the technical rider.  One of the positive aspects of such a project is that cities like Cleveland are looking for symbols of Industry that also represent innovation and creativity.  Industrial robots are both interesting machines themselves, yet to some, they represent the loss of human jobs.  I find the story of such a machine becoming unused, then being donated to a group of hacker artists even more interesting.  But that narrative only goes so far until we actually hack it and unleash the art.  What will it take?  If we can put a dollar figure on it, it has a chance of coming alive; especially since we already know it "works".

[Ross Bochnek]

Given the difficulties of returning the equipment to its original operating state, what are the possibilities of reconfiguring just parts of it to be MUCH easier to power, program, use, and transport?  Otherwise, it sounds easier and safer to just draw inspiration from for a scratchbuild.

[Jonny Hall]

Did it break or something? Last I was involved, it was pretty much fully working -- we had steered it with the teach pendant and written programs for it.

On Mon, Mar 24, 2014 at 10:15 PM, Ross Bochnek <interac...@gmail.com> wrote:

Given the difficulties of returning the equipment to its original operating state, what are the possibilities of reconfiguring just parts of it to be MUCH easier to power, program, use, and transport?  Otherwise, it sounds easier and safer to just draw inspiration from for a scratchbuild.

--

[Matt Hummel]

It works. Just after moving to the new location in the building we lost our access to the power box and it hasn't been bolted down again. Not too mention or method of hooking power to it wasn't very good

[Ross Bochnek]

Maybe a frank Fanuc Robot discussion can be on the agenda for this or next week's Open Hack meeting.  We're in the home stretch for Maker Faire this week, and we'll probably be bagging blinky kits tonight.  But, waiting until next week only gives us a week to write up any proposal and budget for Ingenuity Fest.  It would be great to gather Jonny, Joe, Matt, and I together in person around the Fanuc.

[Joe O'Donnell]

 I'm attending a Big Data event this evening, and can't make the meeting. I will be at LH this Saturday morning for my Data Science meetup.

Yes Ross, extending our recent conversation on the robot to a meeting when we can would be good.

 In my opinion, safety, electrical, transportation and cost issues make the robot unsuitable for a mobile display.

 This is because of the code requirements for 3 phase power, the manufacturer requirement of the robot being bolted to the floor

when used, the manufacturer requirement for a fence to be placed around the robot when in use, the manufacturer requirement for an overhead forklift or crane when moving it, the manufacturer requirement for an industrial scale tie down system when transporting it, and the significant limitations of it's 30 year old control system.

It might be desirable to setup the robot in the space, as opposed to giving it away or scrapping it, but before going on that course, it would be good to consider that:

-Doing that would be a significant decision because significant issues of cost, time, safety, infrastructure, obsolete parts and programming, lack of documentation and difficulty of use are involved.

 The robot looks impressive, but has significant limitations:

-We don't have a hand actuator.

-We don't have adequate documentation to know how to store programs in the hard drive of it's proprietary controller.

-We don't have adequate documentation to know how to communicate with it's proprietary controller over a serial port.

-The robot often stops moving when being pendant controlled and requires a controller reboot. That might be normal, or it might be because of some failing 30 year old obsolete proprietary component that is unlikely to be replaceable. This, and the lack of documentation and ability to store programs, makes entering 'pendant programs' laborious.

-The robot in recent use seemed to fall out of precise movement calibration quickly. This could be another age factor.

-We don't have a way of backing up the software in it's proprietary controller, and thus if we accidentally erase it or make modifications we later don't like, we have no way of restoring it to a working state.

-While researching this almost 2 years ago, I found out it uses obsolete proprietary motor position encoders which others have tried and failed to reverse engineer. That means that making our own controller for it is probably not possible.

-It must be supervised when in use by people who have studied it's manuals on how to operate it safely, following such industry mandatory practices as lockout-tagout.

 All that said, if we were able to meet the manufacturer requirements and other costs for setting it up in our space, we would achieve a significant benefit:

-An impressive looking working robot arm, able to at least move and push objects around.

 That's a significant benefit, and it well may be worth it to put in the significant amount of time and cost to achieve that.

Before deciding on that course, it's useful to look at other possible ways to spend our limited time, money and infrastructure focus on. For example, on working with the open source servo:

http://openservo.com/

The open source servo deals with the modern central problem in robotics, which is how to get better power burst control over the motors. Solving that problem is what enabled SCHAFT robotics to win round 1 of the DARPA humanoid robot contest:

http://spectrum.ieee.org/automaton/robotics/humanoids/schaft-robot-company-bought-by-google-darpa-robotics-challenge-winner

 

---

From: Ross Bochnek <interac...@gmail.com>
To: makersa...@googlegroups.com
Sent: Tuesday, March 25, 2014 8:05 AM
Subject: Re: [makersalliance] Re: Industrial Robot Hacking

Maybe a frank Fanuc Robot discussion can be on the agenda for this or next week's Open Hack meeting.  We're in the home stretch for Maker Faire this week, and we'll probably be bagging blinky kits tonight.  But, waiting until next week only gives us a week to write up any proposal and budget for Ingenuity Fest.  It would be great to gather Jonny, Joe, Matt, and I together in person around the Fanuc.

[Kendall Smith]

Joe,

  Excellent summary of the situation.

  My two cents:

  The current controller is not suitable for our use.  I would look into the feasibility of inserting a more modern controller, while utilizing the existing sensors and servo control circuitry.  I make this suggestion without having a good comfort with the feasibility of this approach.  If using the current sensor and servo control interfaces is not practical, these would also need to be re-made.  Electronics is relatively easy to recreate, the mechanics of this robot are what is irreplaceable.

  I agree that this robot is not suitable for mobile use.  It is too forceful to use in any "open" environment.  However, with a suitable mount and fence we should be able to have fun with it.

  I would like to understand the "loss of calibration" that Joe mentions.  Does anyone understand what is happening?

  If we have a "what do we do with the Fanuc?" discussion on Saturday, I will be there.

 - Kendall

--

Kendall Smith

440-602-2908

KASmit...@gmail.com

[Joe O'Donnell]

 Thanks Kendall. I'll be done with my study group at 12, and other club members often show up around that time also.

The loss of calibration behavior was observed when the robot was being pendant programmed last fall, to move towards

a button on a spray can and press it. The robot would sometimes change were it was pressing by perhaps an inch.

I mentioned that in part to show that the robot isn't currently useful for CNC like purposes.

 If you would like to help create a new encoder and motor control system for it, that would be fantastic. In addition to reusing

the robots mechanical system and motors as you mentioned, we could experiment with SCHAFT like fast torque response.

---

From: Kendall Smith <kasmit...@gmail.com>
To: makersa...@googlegroups.com
Sent: Tuesday, March 25, 2014 11:25 AM

[Ross Bochnek]

[Ross Bochnek]

Many of us will be at Mini Maker Faire on Saturday at CPL downtown.

--
You received this message because you are subscribed to a topic in the Google Groups "Makers' Alliance" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/makersalliance/6cSCTxjeJV8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to makersallianc...@googlegroups.com.

[Matt Hummel]

Joe, when we were doing the whipped cream thing with the bot last year that was entirely done under Manual control with the pendant. I'm not sure if it was losing calibration as much as just the button was held down a little too long. As far add the lock up issue, it would only happen when one of the axis would hit its end stop.
I think some more work with it needs done before we recreate the wheel

[Joe N]

I agee with Matt. More work on the current configuration needs to be done.

[Joe O'Donnell]

 Thanks for the info, and yes, learning whatever we can with the current system and documentation would be great

before working on various potential modern upgrades.

 Here is a video of a test of the robot:

https://www.youtube.com/watch?v=ZzReBB7FF8g&feature=youtu.be

---

From: Matt Hummel <humm...@gmail.com>
To: makersa...@googlegroups.com
Sent: Thursday, March 27, 2014 6:56 AM

Subject: Re: [makersalliance] Re: Industrial Robot Hacking

Joe, when we were doing the whipped cream thing with the bot last year that was entirely done under Manual control with the pendant. I'm not sure if it was losing calibration as much as just the button was held down a little too long. As far add the lock up issue, it would only happen when one of the axis would hit its end stop.
I think some more work with it needs done before we recreate the wheel

--
You received this message because you are subscribed to the Google Groups "Makers' Alliance" group.

To unsubscribe from this group and stop receiving emails from it, send an email to makersalliance+unsub...@googlegroups.com.

[Ross Bochnek]

I recently noticed that the door on our big Fanuc industrial robot arm controller was opened, and THAT'S OK!  Thanks for your interest!

I'd like to talk to whoever was interested enough in it to open it (which probably included clearing some space around it).  Let's learn about this beast.

BTW, whomever is interested in 1980's vintage industrial robot restoration should be interested in our current Seiko Robot Hacking project, in which we are creating a custom controller for our tabletop, 4-axis, blue, Seiko bot, using an Arduino and Raspberry Pi.

[Ross Bochnek]

I checked out one of the Fanuc's motors, and one spec is 60 Volts DC.  

One way to controll the big arm besides using the included controller would be to build a custom one like we're progressing on for the Seiko industrial robot.   A treadmill motor controller would handle that, although I still have to look into how we could control the treadmill controller.  One would would probably be required for each motor, and they seem readily available on eBay starting around $40 each.

[Jonny Hall]

If the goal is to control the arm more easily than with the somewhat clunky but functional programming interface currently available, my suggestions would be either (1) figure out the serial interface to the current controller (it's been years since I've looked at it but I seem to remember there being one; if nothing else perhaps the remote pendant interface could be reverse engineered) or (2) use the existing servo drives and reverse engineer an interface to them.  As I recall we had pretty good documentation on the controller unit, and might as well make the most of what's there rather than spending money to hack something else together.

Of course, if the goal is to learn how to run servos and build something, then have fun!

On Fri, Oct 16, 2015 at 5:43 AM, Ross Bochnek <interac...@gmail.com> wrote:

I checked out one of the Fanuc's motors, and one spec is 60 Volts DC.  

One way to controll the big arm besides using the included controller would be to build a custom one like we're progressing on for the Seiko industrial robot.   A treadmill motor controller would handle that, although I still have to look into how we could control the treadmill controller.  One would would probably be required for each motor, and they seem readily available on eBay starting around $40 each.

--

You received this message because you are subscribed to the Google Groups "Makers' Alliance" group.

To unsubscribe from this group and stop receiving emails from it, send an email to makersallianc...@googlegroups.com.

[Kendall Smith]

Strongly agree with Jonny Hall.  If we want to make the Fanuc move in a useful way, try to work with the existing electronics.  As I recall, this unit was operational when we got it.

 - Kendall Smith

[Nathan Clark]

I agree as well. It may also be useful to check out what is available for interfacing softwarewise in Industrial version if the open sours Robot Operating System  ROS-Industrial  @ rosindustrial.org 

Sent from my Verizon Wireless 4G LTE smartphone

-------- Original message --------
From: Kendall Smith <kasmit...@gmail.com>
Date: 10/16/2015 1:51 PM (GMT-05:00)
To: makersa...@googlegroups.com
Subject: Re: [makersalliance] Re: Industrial Robot Hacking

[Ross Bochnek]

ROS-Industrial sounds amazing!  

I'm already very interested in using ROS for a Roomba project.  I want to try running it in Lubuntu on a Raspberry Pi 2.