First Remap project(s)

[Damian Axford]

HI All

Following our meeting with Adam from Remap the other week, he's found two projects that sound a good match for us:

1) Reading Aid

A reading aid for a lady with age related macular disorder. She has only poor peripheral vision remaining and needs text to be large, sharp and of adjustable foreground/background colour and high contrast. There are commercial magnifiers but most are based on analogue video and are very expensive (way beyond her means). I was thinking about something based on a Raspberry Pi with associated camera and a common wide screen TV . It would need to go on a hand controlled rail to traverse a book or newspaper (we could easily do that part).


2) Robot Arm "Feeder"

Using a low cost robot arm to make a feeder for a lady with cerebral palsy. She lives independently but her hand/arm control is now too poor for her to feed herself. The only commercial offering is the £2,000 "Neater Eater", basically a large mechanical device with fluid damped couplings. She has tried one but it can't be fitted in her kitchen and she didn't find it easy to use.

Something like the cheap plastic robot arm kits at about £30 are not strong enough but something like Aluminium Robot Arm at the bottom of http://www.robot-electronics.co.uk/acatalog/Robot_Kits_Bases.html

or

http://www.conrad-electronic.co.uk/ce/en/product/191516/Arexx-RA1-PRO-Metallic-Robot-Arm;jsessionid=23E117F7A6C92E52C518A59470B9A74B.ASTPCEN28?ref=list

may be suitable (it only has to manage spoon, not a knife and fork!). It would need a stand alone controller which could be programmed to scoop the food up and raise it for eating.

-----------------------------------

Both seem like excellent matches...  would be great if you could let me know appetite to work on either, and any thoughts on what kit we already have between us that could be put together in a proof of concept.

thanks

D

[Damian Axford]

Link to Neater Eater site:

http://www.neater.co.uk/eating-aid

[Damian Axford]

A similar project in Brazil:

https://www.behance.net/gallery/4428639/Eating-Robot-1st-Prototype-(work-in-progress)

[Damian Axford]

and another one in Holland:

http://www.besticinc.com/products/eating-assistive/

[Simon Webb]

Hi Damian

The museum has a servo based robot arm that I would be happy for you to use as a proof of concept if that helps?

Simon

[Damian Axford]

HI Simon

That'd be great...  I've been researching arm designs and servo options, thinking a SCARA arrangement would be ideal, but a prototype arm would certainly help speed things up!  Will pop down next Wed to collect if that's ok?

thanks

D

[Simon Webb]

That's fine Damian, I will leave it on the side bench.

Simon

[Gyrobot]

Hello Spoon, yeeaay!!!!

http://youtu.be/qSV-TrKMH3Q

https://twitter.com/HelloSpoon

http://hellospoonrobot.com/

No comment.

[Lilla Newton]

Hi,

I wanted to invent the feeding robot myself, but happy that someelse did it, so I will have time for other things...

How much does it cost the phone software and the robot? Is the inventor happy to sell the software and the robot manual for a selfbuild?

Do you know who is the inventor and what is his contact?

Please reply ASAP. Thanks

Lillie

07910 139 176

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

[Damian Axford]

awesome, well found - will do some further digging on that...   looks like either SAM or Dynamixel joints, which probably means it'll have similar torque limitations to the other arms in that class (i.e. <100g at full extension)...  would make it ideal for spoon feeding, but no good for lifting a drink or other tasks (e.g. opening the door on the microwave).

In other news, I got a sip/puff controller (with morse code decoder) working on Friday...  gives a nice rich interface for interacting with robot arms, computers, etc  - plus the kids loved it, might even get them to learn morse code! :)

[Damian Axford]

@Simon - forgot to say I had a look at the arm you left out, but reading it's data sheet, it doesn't have enough torque or reach to meet the requirments - figured it's probably not worth the effort to prototype on something that won't be very re-usable.

Have now ordered/received parts for a prototype design, construction/testing underway...

[Jamie Osborne]

If you can output via a standard protocol and interface (bluetooth?), you can probably use it for iPad/iPhone control too (and perhaps Android and others?)

Check out switch control in accessibility, there's a few pages of configuration for such devices.

[Damian Axford]

good call...  that'll be how the commercial units are doing it - accessibility switch + Bluetooth HID device....

this is one of my reference designs:

http://www.orin.com/access/sip_puff/

[Damian Axford]

Another project request - please let me know if anyone is keen and I'll co-ord with Remap:

Amazon Kindle Page Turner

Client request:

"Due to a form of Muscular Dystrophy, it's been a very long time since I've been able to hold up a book to read in bed, a definite problem for a student and self-confessed book addict like myself. I bought a cheap but functioning Kindle from ebay with the hope of being able to work it, but the buttons require too much pressure. Would you be able to attach a remote switch/switches to it?"
 

Prior art:

• http://www.orin.com/access/pagebot/
  
• http://www.ifixit.com/Guide/Kindle+3+Remote+Page+Turn+Switch+Replacement/5749 ?
  

Thoughts:

• Could either go the ifixit route and tap into the internal button connections, or go for a non-invasive servo option (which should make it easier to replicate/adapt for different models of e-reader) - I'd favour the servo route, albeit less elegant
• Buttons needs to be a good size, and require low pressure to activate
• Need to look at how the device is initially turned on, book selected, etc... 

[Jamie Osborne]

Hi Damian,

I like the "ifixit" route as a neater, cheaper, and less complex job. Though a mechanical system that adapts to the reader would be good as a more general solution, but more expensive and cumbersome.

As for activation, the big button suction-cupped to a desk seems a bit awkward, have you considered either:

- pressure switch activated by squeezing a ball - disadvantage is only one "action" per ball.

- optical or proximity switch, moving hand in range activates switch - disadvantage is less tactile, and an "invisible" switch could be confusing or temperamental.

- tilt switch mounted in a hand grip (or anything else comfortable to hold while reading). I think this seems the best and most elegant solution - just tilt hand in either direction to page forward or back. The switches are so small, it could be fitted inside a wrist strap.

Jamie

[Jamie Osborne]

Do you know if the kindle itself can be comfortably held while reading? If so, there's probably room inside for the tilt switch...

Jamie

[Damian Axford]

hey - liking the tilt switch idea...  I discussed progress on this with the assigned engineer last night (Richard).  He's working on a first edition using micro-switches, but would welcome assistance if you'd like to take it on and try the tilt switch idea?  Remap will refund any parts cost, not that it's likely to cost much.  He's also due to visit the lady shortly to understand how she is holding the device (e.g. table stand, etc).

If you'd like to get involved, I can put you in touch with the right people?   Richard is a little swamped right now, so would welcome someone taking this on.

thanks

D

[Jamie Osborne]

If the

Jamie

[Jamie Osborne]

(Doh! Erroneous Send!)

If they have a kindle they don't mind me taking apart, I can take a look to see if a tilt switch can be squeezed in there.

Jamie

On 12 Aug 2014, at 13:56, Damian Axford <dam...@axford.me.uk> wrote:

[Damian Axford]

they most certainly do... the lady in question bought one 2nd hand specifically to be taken apart and modified...  will send you some contact details via email :)

[Jamie Osborne]

Excellent, it's a shame amazon don't think to add such a simple alternative. Do any of the keyboard buttons (cursor arrows?) allow page turning*, or are they equally hard to press (I haven't used a kindle much).

* I'd assume not as this would be the obvious solution without any effort required! ;)

Jamie

[Damian Axford]

I don't believe so... but similarly, I've only used one briefly a long time ago...

[Gyrobot]

Another alternative for controlling devices :

http://www.instructables.com/id/Muscle-EMG-Sensor-for-a-Microcontroller/

[Damian Axford]

cool... and for the lazy (like me), their latest kits are $49 with a teeny little controller board:

http://www.advancertechnologies.com/p/muscle-sensor-v3.html

very neat - that goes on the "to experiment with" list...    we could always buy one with the hackspace kitty to play with?  perhaps put together a complete myoelectric demo arm? 

[Damian Axford]

Robot arm update (affectionately named AFRo)....   for anyone who wants to follow the design process, I've configured my design notes to auto-publish at the following location:

https://docs.google.com/document/d/1HDJsVby1FxGu6IQ2XBOvq7dboFC9hibPxshcKR86C0M/pub

It's quite a long doc already (20+ pages), but should be interesting reading.  I've also published the WIP CAD files, source code, etc to github here:

https://github.com/Axford/AFRo

Although the man-cave has taken up most of my time, I've got some basic motion code for the stepper and servos working (via interrupts).  Should make a lot more progress in September, once man cave is complete.

:D

[Gyrobot]

Enjoyed the read, maybe myoelectric option in "Future Exploration"? using the little sensors in the link above?

[Damian Axford]

thanks and good idea - will add it on

[Damian Axford]

Video of the mk1 arm, running a basic motion test.... this was from last week:

http://youtu.be/KAB5SaC1I_I?list=UUeu1FeCzFXfNYYMqFcKwptg

Currently hacking Teacup firmware to control 5-axes and extending it to support the Dynamixel/Robobuilder servos.  Have 3-axes Z + Shoulder + Elbow working ok as of last night :)    Might bring the arm to oggcamp if it's in a demo'able state

[Jamie Osborne]

Looking good! And that'd be an awesome idea!! :)

Jamie

[Damian Axford]

cheers mate -   here's the 1st four axes working together:

http://youtu.be/d4N9LCZNLSA?list=UUeu1FeCzFXfNYYMqFcKwptg

The arm is being controlled by a little host app (written in Processing), which can send joint positions individually or execute a complete script (as in the video)

Slightly harder job of extending the firmware to handle 5-axes, and integrate the Robobuilder servo that controls the spoon tilt angle...   might get that done tonight ready for oggcamp

[Robert Longbottom]

That looks pretty neat, the motion looks nice and smooth as well.  Will it be variable speed?  So if it you are starving you can put in on fast but it might drop a few peas in the process :-)

--

[Damian Axford]

lol - yes, it can do variable speed...    :)      Uses the feedrate functionality already in the firmware, along with acceleration and feedrate limits.  Works really well.

Unfortunately the motion is not as smooth as I'd like (looks better in the video than it is) - the servos are no way near as "good" as I was expecting.  There is quite a bit of tremor at the spoon tip when doing fast vertical moves or slow horizontal moves.  I think I've pushed the limit of the current mechanical design, so mk2 won't be far away - this is roughly what I have in mind:

Red bits are motors - replacing the shoulder and elbow servos with steppers, and the threadless-ballscrew Z axis with a belt drive (and sliding counterweight).

Mk2 should be faster, smoother, stiffer and possibly cheaper    :)    (in theory)

[Robert Longbottom]

That's look pretty smart.  What software are you using for designing the models?

Will V3 have a Gyrobot hand on the end for finger buffets? :-)

[Jamie Osborne]

Ha!

Looks plenty good enough to show off tomorrow I think?

Jamie

[Damian Axford]

@Rob - thanks.  The printable designs are in OpenSCAD, but I tend to do concept designs (like the mk2 one above) in Sketchup.     On the subject of gyrobot hands, and other things awesome, Jamie and I got talking about doing some PR stunts whilst at oggcamp - will come along Wed to discuss more, but could lead to some fun group projects.....   for this one, I'm thinking giant flexy hand...  animated with motors'n'stuff and stuck on a trolley to take through the town.  Like the giant lottery hand - "It's You"

[Damian Axford]

Quick update on the arm...  the mk2 design is very nearly finished, just a few parts and some wiring to go before it's back to s/w dev.  I'm very pleased with how fast this version has come together - been just over a week since I sketched out a mk2 design - here's the before/after:

just shows how fast you can get things done with rapid prototyping :)

[Damian Axford]

and this is the current state of the OpenSCAD model - you can see some of the parts still to be printed in here (and a little modelling error with the stepper at the bottom :) )

[Robert Longbottom]

That's pretty impressive for an openscad model!

On 13 October 2014 14:45:19 Damian Axford <dam...@axford.me.uk> wrote:

and this is the current state of the OpenSCAD model - you can see some of the parts still to be printed in here (and a little modelling error with the stepper at the bottom :) )

[Damian Axford]

thanks - i do like my OpenSCAD :)

that was going to be my other talk at oggcamp - "extreme OpensCAD"

[Jamie Osborne]

I was just thinking to suggest another OpenSCAD talk (perhaps because you already said this at oggcamp!) - I could do with learning the basics first before anything extreme though!

Jamie

[Damian Axford]

I'd be happy to do one, perhaps in combo with Jess (to re-use her intro talk)?

the "extreme" bit kicks in when you want to get into multi-part models, or complete machine design....  then you need some different techniques to work efficiently.

off the top of my head, it'd break it down like this (pre-req is a basic understanding of 3D modelling terminology/concepts):

1) Intro to OpenSCAD - language basics, exporting to STL (this is pretty much covered by Jess' talk)

2) Common libraries (utils and vitamins) - an overview of the most useful libraries out there

3) Multi-part models - coding best practise, modularity, using "attach", animations

4) Machine design - concepts, file structure, bulk generating STL, generating BOMs

Each section would be 1-2 hrs long

[Robert Longbottom]

I'd be quite interested to learn some more advanced openscad.  Though I'm doing more stuff in freecad these days.

[Jamie Osborne]

Yeah, that looks good! I'd be interested in a recap of the talk Jess did (or equivelent) and some more advanced stuff. 

Jamie

[Damian Axford]

ok...  so I always like to have a real project to work on to learn stuff quickly.  

Any suggestions for a "group" exercise in OpenSCAD - that could form the underlying theme for a series of workshops?   

A good recent example would be the brackets Rob made for the teletype paper roll.  A basic model would include a simple representation of the teletype outer case, the paper roll and then the brackets themselves.

Of course, the brackets are done, so I'd rather we work towards something new/useful...

If we can generate a theme tomorrow, then I'm happy to do an intro talk on Wed evening - limit it to an hour and make it very hands on.  We can then build on it in future weeks...  

[Jamie Osborne]

Sounds good. A very simple thing I'd like to model (hopefully to print at the Hackspace), would be a Pi bracket/sled for the teletype. The Pi is a rev1 model B, so no "mounting" holes - though I could swap it for my rev2 if really needed.

Jamie

[Damian Axford]

ok - great, a PI case will get us through the first two sessions, and we could always model the teletype casing for parts 3+4

alternatively - we could switch to modelling this turtlebot idea

[Jamie Osborne]

The Pi "case" I was thinking doesn't even need a lid - so we could do that and the logobot* ?

* I think I like this name better - as turtle == little robot that is programmed with logo!  ;)

Jamie

[Jess Robinson]

Ah, hadn't realised this thread had mutated into an OpenSCAD lesson thread
;)

That sounds like a good idea, I was hoping we could do something similar
(but more of a short intro) for make-shift-do as well.

I can redo mine if you like, though I'd need to revise it as parts were a
little confusing (the overlappin/difference example needs fixing).

I'd like to attend some of those last ones :)

Jess

On Mon, 13 Oct 2014, Damian Axford wrote:

> I'd be happy to do one, perhaps in combo with Jess (to re-use her intro
> talk)?
>
> the "extreme" bit kicks in when you want to get into multi-part models, or
> complete machine design.... then you need some different techniques to
> work efficiently.
>
> off the top of my head, it'd break it down like this (pre-req is a basic
> understanding of 3D modelling terminology/concepts):
>
> 1) Intro to OpenSCAD - language basics, exporting to STL (this is pretty
> much covered by Jess' talk)
> 2) Common libraries (utils and vitamins) - an overview of the most useful
> libraries out there
> 3) Multi-part models - coding best practise, modularity, using "attach",
> animations
> 4) Machine design - concepts, file structure, bulk generating STL,
> generating BOMs
>
> Each section would be 1-2 hrs long
>
> On Monday, 13 October 2014 16:23:13 UTC+1, jmeosbn wrote:
>>
>> I was just thinking to suggest another OpenSCAD talk (perhaps because you
>> already said this at oggcamp!) - I could do with learning the basics first
>> before anything extreme though!
>>
>> Jamie
>>

>> On 13 Oct 2014, at 15:56, Damian Axford <dam...@axford.me.uk <javascript:>>

>> wrote:
>>
>> thanks - i do like my OpenSCAD :)
>>
>> that was going to be my other talk at oggcamp - "extreme OpensCAD"
>>
>> On Monday, 13 October 2014 15:01:21 UTC+1, Robert Longbottom wrote:
>>>
>>> That's pretty impressive for an openscad model!
>>>
>>> On 13 October 2014 14:45:19 Damian Axford <dam...@axford.me.uk> wrote:
>>>
>>>> and this is the current state of the OpenSCAD model - you can see some
>>>> of the parts still to be printed in here (and a little modelling error with
>>>> the stepper at the bottom :) )
>>>>
>>>>

>>>> <https://lh5.googleusercontent.com/-dck4Tg5fDjY/VDvXSJKdqAI/AAAAAAAAAVo/PkCu9_bZ2Yk/s1600/mk2%2Bwip1.png>

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

>> email to swindon-hacksp...@googlegroups.com <javascript:>.