These are chat archives for nightscout/intend-to-bolus

7th
Oct 2015
Jason Calabrese
@jasoncalabrese
Oct 07 2015 00:35
like the --extra @bewest, no more private repo
Ben West
@bewest
Oct 07 2015 00:35
:-D
Jason Calabrese
@jasoncalabrese
Oct 07 2015 00:35
would be cool to have a registry of public operational logs
repos
screwed up my pi's network config, and haven't had the time to fix it
but at least I got the battery charged
Pete Schwamb
@ps2
Oct 07 2015 01:35
@bewest how do you know there’s a command to read the display?
Ben West
@bewest
Oct 07 2015 03:01
nice... I fixed a bug... my entire loop runs in under a minute now
Dana Lewis
@danamlewis
Oct 07 2015 03:01
:)
Scott Leibrand
@scottleibrand
Oct 07 2015 03:01
Sweet! What was the bug delaying?
Ben West
@bewest
Oct 07 2015 03:02
hmmm, there were issues with handling model type and fetching the model too often
there were also issues with doing wrong thing with expires
but separate, fixed separately
I did something today at some point which made it much much worse
and noticed in the rileylink sniffing it seemed to be asking for model little more often than I thought it should
fixed it up a bit, now it's nice and fast
Scott Leibrand
@scottleibrand
Oct 07 2015 03:09
Nice. Will have to pull your latest and test it when I get home.
Pete Schwamb
@ps2
Oct 07 2015 04:14
Wow, I just disassembled an omnipod pod. It’s like looking at alien technology. Lots of stuff I haven’t seen before.
How can they sell these for $35 apiece?
Pete Schwamb
@ps2
Oct 07 2015 04:19
it has muscle wires for a motor.
Jason Calabrese
@jasoncalabrese
Oct 07 2015 05:03
$ sudo bin/fix-dead-carelink.sh
Power-cycling USB to fix dead Carelink stick
bin/fix-dead-carelink.sh: 18: bin/fix-dead-carelink.sh: cannot create /sys/devices/platform/bcm2708_usb/buspower: Directory nonexistent
bin/fix-dead-carelink.sh: 20: bin/fix-dead-carelink.sh: cannot create /sys/devices/platform/bcm2708_usb/buspower: Directory nonexistent
guess bcm2708_usb is device specific?
Holger Sanft
@holgi13
Oct 07 2015 05:15
Anyone flight with xDRIP (not the complete Rig)? Starting and landing with xDRIP ON is possible? My wife and me, go next week to a trip to London.
Dana Lewis
@danamlewis
Oct 07 2015 05:22
@holgi13 Might want to also post this in the xDrip channel, too, not sure how many here are using it
Holger Sanft
@holgi13
Oct 07 2015 05:31
@scottleibrand openaps work the last three nights fine. But high glucose entry to the night needs hours to landing in target area. Same thing in the middle of this night, pump out of Insulin, change catheter, restart system with 250 mg/dl in a test, no bolus. Target area landing in two a half hour.
Dana Lewis
@danamlewis
Oct 07 2015 05:33
@holgi13 Do you think it's not bringing down quickly enough?
Holger Sanft
@holgi13
Oct 07 2015 05:34
@danamlewis good idea or I need to go back to cable
@danamlewis yes I think so. It's possible to bringing down faster only in high range maybe?
Chris Hannemann
@channemann
Oct 07 2015 05:45
@jasoncalabrese I know that @eszcloud had the same issue
Jason Calabrese
@jasoncalabrese
Oct 07 2015 05:47
pulling the stick works, but it is a reliability issue
Ben West
@bewest
Oct 07 2015 05:52
that script has nothing to do with the carelink script
stick, too
Holger Sanft
@holgi13
Oct 07 2015 06:10
@danamlewis you are right it is safety, but for the future I think, a idea is to split the regulation in three parts? High range - target range - low range?
Scott Leibrand
@scottleibrand
Oct 07 2015 12:32
@holgi13 what is your max_basal and max_iob? How long does it high_temp you at the max rate when you start super high?
the algorithm is really designed to high-temp you as you're rising, not after you get high.
Also, what branch of openaps-js are you using? The newer dev and js-only branches have better logic for continuing to high-temp when BG is falling, but not as much as expected. If you're still on master and want it to be more aggressive, you should try switching.
@bewest that script does power-cycle the carelink stick (along with all USB devices), but apparently only on the older Ali
Stupid gitter app. On the older Pi B.
It does nothing on the Pi 2 or other hardware.
Scott Leibrand
@scottleibrand
Oct 07 2015 12:38
Need to figure out the correct /sys/devices path for doing the same on Pi 2
Ali Mazaheri
@amazaheri
Oct 07 2015 14:20
lol
Scott Leibrand
@scottleibrand
Oct 07 2015 14:51
:)
Ben West
@bewest
Oct 07 2015 15:12
@scottleibrand that script has no effect on carelink script
Ben West
@bewest
Oct 07 2015 15:19
I've renamed it to be less confusing
Scott Leibrand
@scottleibrand
Oct 07 2015 15:19
How do you know that?
It reboots all USB devices on the Pi
Which empirically does fix the issue
I've watched the power get cycled to both wifi and dexcom when it runs, and presumably it cycles the carelink's power as well, because it goes from unable-to-scan to being fine.
I don't understand why you think that's unpossible.
Ben West
@bewest
Oct 07 2015 15:22
that script only does one thing
it only has an effect when a particular wifi dongle is used
if that wifi dongle is not used, the script has zero effect
Scott Leibrand
@scottleibrand
Oct 07 2015 15:23
How come it kills power to the Dexcom then?
and how come it fixed the dead carelink stick?
i don't know what it "should" do, but what it actually does is powercycle all USB devices on the original Pi B
when I get home I could test whether it still does that when using a different wifi adapter...
and/or whether it works on the Pi 2 with the same wifi adapter we use on the old Pi B
Ben West
@bewest
Oct 07 2015 15:32
reading the script, I would suspect that the hardware or the kernel module is extremely buggy
that script only does one thing, if other things are happening, that might actually be an indication of the problem rather than the solution
ah, ok on fourth reading...
bcm2708_usb is indeed the entire usb stack on rpi
Scott Leibrand
@scottleibrand
Oct 07 2015 15:39
Do you know offhand what it is on Pi 2?
Oliver Schumacher
@oschumac
Oct 07 2015 15:41
@scottleibrand. On my pi the script has killed the wlan. Everytime i had to reboot. So i switched to reboot if decocare scan returns a false.
Ben West
@bewest
Oct 07 2015 15:42
that script still has nothing to do with the stick ;-) I keep trying to find information about bcm2708 but it's all folk-lore on internets
it appears it controls powers to entire usb stack and does some things on pi, and that people are using it to reset their wifi/lan configs
Scott Leibrand
@scottleibrand
Oct 07 2015 15:43
@oschumac interesting, my wlan comes right back.
Oliver Schumacher
@oschumac
Oct 07 2015 15:45
Reboot is okay for me. need 30 sec. than be online again. It only sucks if you are troubleshooting. And being kicked out from ssh session.
Scott Leibrand
@scottleibrand
Oct 07 2015 15:46
Yeah, that was why I didn't want to reboot. For me USB cycle is just a 2-second blip.
Ben West
@bewest
Oct 07 2015 15:49
crap my pump is dying
motor error frequently
crap
need to find replacement stat
Scott Leibrand
@scottleibrand
Oct 07 2015 15:51
What pump(s) do you have?
Ben West
@bewest
Oct 07 2015 15:51
I have a 522 that is about to stop working altogether
european 722,
hmmm guess that's it
used to have some others but they turned to dust
Scott Leibrand
@scottleibrand
Oct 07 2015 15:53
We have one dev/backup 515, and an emergency-backup 523 that can't loop.
Ben West
@bewest
Oct 07 2015 16:43
I can probably find emergency backup locally if really pressed
I have a 722 that is european... so I have one european RF stick, but no 7xx supplies
Ben West
@bewest
Oct 07 2015 16:58
wow, I love medwow.com
gotta love those people with gestational diabetes who bought a pump and tried it for a few weeks and put it back in the box
Dana Lewis
@danamlewis
Oct 07 2015 17:00
Find some good prices?
Ben West
@bewest
Oct 07 2015 17:03
yeah
and, no concern about quality, which is bigger deal
when someone says "yeah, I had gestational diabetes and tried it for a few weeks, been in box since then..."
it's basically mint condition...
lol
Dana Lewis
@danamlewis
Oct 07 2015 17:04
:)
Ben West
@bewest
Oct 07 2015 17:04
i've seen some in the past that looked like they use their teeth or claws or something to open the battery, reservoir, etc...
I'm fairly rough on my devices... but it's like how did you even...
the caps are hand tight, not machine tight...
one had almost totally stripped the groove on the battery cap
Oliver Schumacher
@oschumac
Oct 07 2015 17:07
@bewest i use 1.8 ml ampullas in a 722 working fine no props. 3.6 in a 5xx won't work.
Ben West
@bewest
Oct 07 2015 17:07
what now?
what's amupllas?
canula?
is that 30/45 degree insert?
I have one pump that doesn't seem to fit any of my reservoirs
it's a weird shape or something
Oliver Schumacher
@oschumac
Oct 07 2015 17:10
Maybe i am wrong i bought insulin set for a 512. Then got a 722 pump in addition. The set's for 512 are working in a 722.
Ben West
@bewest
Oct 07 2015 17:10
yeah
Dana Lewis
@danamlewis
Oct 07 2015 17:17
I think ampulla is reservoir?
1.8 or 180u is what goes in a 5xx, so the 3whatever would be the bigger reservoir for the 7xx
Ben West
@bewest
Oct 07 2015 17:18
ah, that makes sense
I have a weird 554 that won't take any reservoir
I think someone put glue in the bottom or something
Dana Lewis
@danamlewis
Oct 07 2015 17:20
We PWDs are kinda crazy :) sometimes
Oliver Schumacher
@oschumac
Oct 07 2015 17:22
reservoir yes that what i trying to to say.
Dana Lewis
@danamlewis
Oct 07 2015 17:31
@bewest Next generation pump clips for Medtronic will have something on the clip to help with battery cap. Otherwise lots of people use coins, which is probably why so many are in rough shape
Ben West
@bewest
Oct 07 2015 17:33
here's a pretty pink one
kind of want a pink one
Ben West
@bewest
Oct 07 2015 17:40
oy, that was dizzying shopping spree
that site is really great
lot of 722's at ~$1k range, give or take a few hundred
some of them are in mint condition
lot of "exotic" ones too, 511, 510, 512, 515, even a 508
Dana Lewis
@danamlewis
Oct 07 2015 17:46
Wow
Matthias Granberry
@mgranberry
Oct 07 2015 17:56
I had a 508 once
Ben West
@bewest
Oct 07 2015 18:13
crap my pump is dead
third motor error today
and high bgs
gah
hmmm
Paul Andrel
@stavlor
Oct 07 2015 18:14
@bewest ouch
Dana Lewis
@danamlewis
Oct 07 2015 18:17
@bewest Check your Fbk thread, 2 people with pumps. Not sure how fast they could get them to you.
Ben West
@bewest
Oct 07 2015 18:18
I'm going to a diabetes thing tonight
if I make it that long, hehehe
Matthias Granberry
@mgranberry
Oct 07 2015 18:19
I don't feel comfortable unless I have a bag of syringes w/in a 5-minute walk of me at all times. This stuff never seems to "just work."
Dana Lewis
@danamlewis
Oct 07 2015 18:22
@mgranberry Agreed. Or if not a full bag, at least a few.
Also cause you can then refill reservoirs more easily
Matthias Granberry
@mgranberry
Oct 07 2015 18:25
I actually travel with basal insulin too. It's just easier not having to worry about a backup plan. And that way I don't have to explain to anyone that no, I can't just wait until morning etc.
Paul Andrel
@stavlor
Oct 07 2015 18:26
@danamlewis true they can be refilled pretty easily provided you have one of the plastic handles handy to screw in to them
Matthias Granberry
@mgranberry
Oct 07 2015 18:27
I bet pressure will push the plunger back out.
Ben West
@bewest
Oct 07 2015 18:28
various ways of dealing with pressure
Matthias Granberry
@mgranberry
Oct 07 2015 18:28
I meant the plunger in the reservoir. I doubt you need the handle if you inject insulin directly into the reservoir.
Paul Andrel
@stavlor
Oct 07 2015 18:29
@mgranberry true that would probably work quite well
Peter Miller
@ochenmiller
Oct 07 2015 18:33
@bewest were you able to actually complete a purchase via medwow? I tried to initiate several purchases there, and never heard back from anyone. Just the occasional notice that a listing had been taken down. Will be very interested to hear how it goes.
though, maybe it was because I was only looking at the $400 pumps...
we need to develop the intend-to-bolus "strategic medronic pump reserve"
Scott Leibrand
@scottleibrand
Oct 07 2015 18:35
lol true
the periodic call for pumps on CGM in the Cloud still seems quite effective...
Peter Miller
@ochenmiller
Oct 07 2015 18:36
I'd be quite willing to add $$ to an account that purchased pumps prospectively
pool our resources and share the risk management of being without a connectable pump.
Ben West
@bewest
Oct 07 2015 18:42
right
yes, I have one confirmed offer from FB, most likely incoming
very grateful
I purchased one on medwow, and quoted 4 or 5 more
I would have bought several, but medwow made me go through the quote process
but it'd be nice to pool/team up for a reserve
Peter Miller
@ochenmiller
Oct 07 2015 19:44
I just found a 723 pump on Craig's list w 2.3A software. They are sking $1,000 (negotiable). Does 2.3a software work? Is $1k too much?
Ben West
@bewest
Oct 07 2015 19:45
23 has limited use... no temp basal :-(
err, no remote temp basal
Peter Miller
@ochenmiller
Oct 07 2015 19:45
Ah, thanks!
Ben West
@bewest
Oct 07 2015 19:46
it's really all about the 22's
Peter Miller
@ochenmiller
Oct 07 2015 19:46
Got it!
Pete Schwamb
@ps2
Oct 07 2015 19:50
I see a couple 722’s in the $600 range on craigslist here in the mpls area.
Ben West
@bewest
Oct 07 2015 19:50
craigslist can be "too local"
browse through medwow to see conditions/prices there to get a good idea
craigslist should be a bit cheaper unless it's unused
Pete Schwamb
@ps2
Oct 07 2015 19:52
I just successfully got 2 omnipod PDMs with 25 pods for $400 from north carolina, so craigslist can, at least in my experience, be used for non-local transactions.
n=1 :)
Ben West
@bewest
Oct 07 2015 19:53
how's RE going on that one?
you will make lot of people very happy and a few people very upset
Scott Leibrand
@scottleibrand
Oct 07 2015 19:58
Heh
Pete Schwamb
@ps2
Oct 07 2015 19:59
Well, I’m trying to evaluate whether it’d be an option for us; otherwise I’m not sure it’s worth the time for me to invest; I’m already over-subscribed in that area.
Since I’ve been mentioning looking into ominpods, I’ve been hearing stories of failure to deliver insulin.
Unexplained highs, and high failure rates.
Ben West
@bewest
Oct 07 2015 20:03
mmm
I'm about to have a broken, but booting pump...
it'll be easier to open the end cap on this one, hehe
Pete Schwamb
@ps2
Oct 07 2015 20:05
The guts of the pods are fascinating. At least to me. :)
Ben West
@bewest
Oct 07 2015 20:05
yes, I should open one
think I saw one someone smashed
that's another thing, apparently when they go bad, only way to turn it off is smash it
Pete Schwamb
@ps2
Oct 07 2015 20:05
hah, really?
Ben West
@bewest
Oct 07 2015 20:06
yeah, apparently it's so common, people don't even mention it
it's like talking about tubing on medtronic
Ben West
@bewest
Oct 07 2015 20:07
I'm thinking new open source pump is needed somehow
Pete Schwamb
@ps2
Oct 07 2015 20:07
In this picture you can see the muscle-wire that is the pump’s motor: http://www.circuitben.net/node/371
Ben West
@bewest
Oct 07 2015 20:08
nice
is that a copper wire?
wrapped around something
Pete Schwamb
@ps2
Oct 07 2015 20:09
The silver bits extending from the rest of the plastic base are conductive, and form part of the circuitry.
No, it’s muscle wire, that contracts when voltage is applied.
Ben West
@bewest
Oct 07 2015 20:09
ah, ok
Pete Schwamb
@ps2
Oct 07 2015 20:09
There is no electromagnetic traditional motor in there at all.
Just a wire that flexes and advances the screw that pushes on the reservoir.
Ben West
@bewest
Oct 07 2015 20:10
but it is electromagnetic, same way as a motor?
it's not using magnetism at all?
Pete Schwamb
@ps2
Oct 07 2015 20:10
correct
no electromagnetic stuff at all.
The wire is Flexinol brand.
nickel-titanium. was a lot stronger than I expected, based on how thin it was (finer than a hair)
Ben West
@bewest
Oct 07 2015 20:14
wow
the asante pumps are lot of fun
modular design
the motor is very simple, elegant, mechanical design
threaded plunger is on one way trip, every step locks mechanically
the parts all around break before it goes backwards
lot of is simple mechanical solution
did not have controller to take apart though
Pete Schwamb
@ps2
Oct 07 2015 20:18
the omnipod has a gear latch as well, to make sure it can’t go backwards. the musclewire flexes, and the gear rotates, and the latch advances a notch, and locks so when the wire de-flexes the gear stays advanced.
Ben West
@bewest
Oct 07 2015 20:19
nifty
so is musclewire also flexing one way, or it unflexes? is there a lifetime to the number of flexing moves it can do?
and more interestingly, how might it compare to motor
Pete Schwamb
@ps2
Oct 07 2015 20:21
no, it just flexes for a moment, then unflexes.
Not sure about lifetime.
Scott Leibrand
@scottleibrand
Oct 07 2015 20:23
Is it somehow designed so that the musclewire gets to start each flex from the same length?
Eric
@ecc1
Oct 07 2015 20:24
I'd like to see an open protocol between "headless" insulin pump and controller. Pump would do basal and boluses, controller everything else. Secure comms over BLE, etc. So controller could be a phone app or dedicated device; pumps could be interchangeable.
Ben West
@bewest
Oct 07 2015 20:24
we'll have to build it :-)
Pete Schwamb
@ps2
Oct 07 2015 20:24
@scottleibrand Yes, after flex, it goes back to it’s original length.
Peter Miller
@ochenmiller
Oct 07 2015 20:25
One could save a lot of space on a cartridge model pump if it didn't have a display. Yet, "maker" style manufacturing seems plausible. Wonder if the stepper motors that medtronic or other manufacturers use are available from electonics suppliers?
Ben West
@bewest
Oct 07 2015 20:25
I was thinking use e-ink display
ditch lcd, etc...
Peter Miller
@ochenmiller
Oct 07 2015 20:26
3d print the case
Ben West
@bewest
Oct 07 2015 20:26
yeah, been thinking about maker style, arduino style building blocks
and eventually the blocks get better and better
Peter Miller
@ochenmiller
Oct 07 2015 20:27
design it to use readily available cartridges from medtronic or other
Ben West
@bewest
Oct 07 2015 20:28
there's a new infusion set that JDRF is claiming credit for
Peter Miller
@ochenmiller
Oct 07 2015 20:28
then all the openaps electronic could live inside too
Ben West
@bewest
Oct 07 2015 20:28
supposed to be pretty interoperable
Scott Leibrand
@scottleibrand
Oct 07 2015 20:28
Seems like building a DIY device to directly administer insulin would be extremely dangerous.
Ben West
@bewest
Oct 07 2015 20:28
they say the same thing about everything else we're doing, too ;-)
at this point we have no suitable pumps on the market
Pete Schwamb
@ps2
Oct 07 2015 20:28
Apparently the lifetime of muscle wire is dependent on how hard it’s used. They have a factor they call “memory strain”.
"If Flexinol® actuator wire is used in the appropriate conditions then obtaining repeatable motion from
the wire for tens of millions of cycles is reasonable. If higher stresses or strains are imposed, then the
memory strain is likely to slowly decrease and good motion may be obtain for only hundreds or a few
thousands of cycles. "
Ben West
@bewest
Oct 07 2015 20:30
fascinating
Scott Leibrand
@scottleibrand
Oct 07 2015 20:32
I'm not so worried about perceived danger so much as actual danger. I understand how to validate whether we're issuing the right dosing commands to a device. I don't know how we'd ensure that with fallible mechanical parts...
Ben West
@bewest
Oct 07 2015 20:33
right, that's going to be the tricky bit, I certainly don't know how to do all that
Pete Schwamb
@ps2
Oct 07 2015 20:34
And based on how unreliable 3d printing is… :)
Getting good feedback from motor advancement or lack thereof is critical.
Scott Leibrand
@scottleibrand
Oct 07 2015 20:34
exactly.
Ben West
@bewest
Oct 07 2015 20:35
yeah, making the motor trustworthy is interesting challenge
trustworthiness of motor might not depend on the case at all though
Pete Schwamb
@ps2
Oct 07 2015 20:36
I think medtronic pumps have two redundant methods of measuring motor advancement.
The measurement in the omnipod looked like it only would give a single tick after many clicks; too coarse of a resolution.
I might be misreading how it works, though.
Ben West
@bewest
Oct 07 2015 20:50
most of the big vendors like freescale, TI, etc actually have reference designs for medical pumps, "connected glucose meters" and other things
Peter Miller
@ochenmiller
Oct 07 2015 20:55
I love the clockwork aspect of the omnipod. Makes me think one could have a wind up pump!
Frank Best
@efbest
Oct 07 2015 21:12
There was a windig Pump Byte Becton-Dickinson Mann years ago!
windup
Matthias Granberry
@mgranberry
Oct 07 2015 21:23
@bewest it's a nickel-titanium alloy that returns to a programmed shape when heated. You can make linear actuators by stretching it out and making it contract.
Ben West
@bewest
Oct 07 2015 21:24
yeah, been reading up on it
Matthias Granberry
@mgranberry
Oct 07 2015 21:25
it's an interesting idea to use it, but I suspect it eats batteries. I wonder if part of the reason they use a custom IC is for higher current handling capabilities
Ben West
@bewest
Oct 07 2015 21:25
the craziest rig I've seen was basically syringe on a stick that had some kind of ticker that could do basal in italy from 80's
I like a simple motor train like asante has
maybe come up with simple dumb motor train assembly
Matthias Granberry
@mgranberry
Oct 07 2015 21:26
some of the type-2 basal delivery devices on the market now are basically a spring with a slow piston. The idea of using that it terrifying, but I guess it's different when you're only partly on manual.
Ben West
@bewest
Oct 07 2015 21:26
its not a medical device without connecting it to controller somehow
Matthias Granberry
@mgranberry
Oct 07 2015 21:27
I'd make it with a reverse Luer lock (adapters are available) to make a point of it. There's no point in inviting involvement from regulatory agencies.
Ben West
@bewest
Oct 07 2015 21:30
dunno, the way asante worked was you could prime the motor train with standard pen-form vials of humalog
but it's still not a medical device
it can't do anything until you arm it with the controller
Matthias Granberry
@mgranberry
Oct 07 2015 21:31
yeah. The Asante device looked really exciting mechanically. Very simple, but effective. I hope Bigfoot makes it to the market soon.
Matthias Granberry
@mgranberry
Oct 07 2015 21:39
I also like that there is so little opportunity for user error. The only downside is that penfills are somewhat expensive to get from insurance.
Did they require special infusion sets? Something has to attach to the septum. The big problem I see with a DIY device isn't accurate delivery, but maintaining sterility and managing consumables.
Ben West
@bewest
Oct 07 2015 21:41
bigfoot is saying 2 - 3 years
Peter Miller
@ochenmiller
Oct 07 2015 21:41
I think you'd want to work with consumables from an established manufacturer
Matthias Granberry
@mgranberry
Oct 07 2015 21:43
of course, but they aren't likely to want to get involved with a DIY insulin pump.
Ben West
@bewest
Oct 07 2015 21:44
I think DIY is maybe not a handle that's serving us well
Peter Miller
@ochenmiller
Oct 07 2015 21:44
No, but the supplies are readily available
Ben West
@bewest
Oct 07 2015 21:44
a lot of focus is put on "tweaking" things and getting additional features
Pete Schwamb
@ps2
Oct 07 2015 21:44
There are 4 1.5v watch style batteries in the pod.
Ben West
@bewest
Oct 07 2015 21:45
in the event where some magic vendor produces some perfect pump, you'd still want to run diagnostics on it, still want to choose which apps/devices can configure it, get history, etc...
Matthias Granberry
@mgranberry
Oct 07 2015 21:45
Nitinol is such a crazy design choice, but I guess it works for them.
Pete Schwamb
@ps2
Oct 07 2015 21:48
@mgranberry I think it’s a cost issue; no way they can put a motor in every pod and keep it under $40.
Matthias Granberry
@mgranberry
Oct 07 2015 21:48
Was the motor not in the Asante cartridge?
Then again, their design might have been related to their demise.
Ben West
@bewest
Oct 07 2015 22:07
pump market is tough
you pay lot more money for extra supplies and you get relatively same level of care, according to the numbers
Matthias Granberry
@mgranberry
Oct 07 2015 22:08
@bewest do you know what regulatory hurdles there would be involved in developing and getting pump legos approved? Would it just be a matter of 3d printing a reference design, getting electronics going, and kickstartering a group run with some injection molded gears?
Ben West
@bewest
Oct 07 2015 22:08
so pump's have not proven their value
Matthias Granberry
@mgranberry
Oct 07 2015 22:08
you mean haven't?
Ben West
@bewest
Oct 07 2015 22:09
main thing with regulatory is avoiding surprises
key thing is avoiding distributing a "medical device"
so would approach them with each block and ask "is this a medical device"
I dunno, my vision for all this has been we'd do all the same stuff normal vendor does
Matthias Granberry
@mgranberry
Oct 07 2015 22:11
but surely they'd look at the firmware and eventually say "yes"
Ben West
@bewest
Oct 07 2015 22:11
just out in the open
nah, they don't really do that it turns out
you have to prove people review your firmware
so that's where a lot of the cost comes in
Matthias Granberry
@mgranberry
Oct 07 2015 22:11
What costs are involved? Is it primarily labor to prepare the FDA review?
Ben West
@bewest
Oct 07 2015 22:11
they hire "outside" reviewers
there's a cost for listing your firm
and hmmm
there's some other costs associated with big filings
nightscout foundation is already listed :-)
Matthias Granberry
@mgranberry
Oct 07 2015 22:12
ah, but with a git PR style development that's largely proven, right?
Ben West
@bewest
Oct 07 2015 22:12
exactly ;-)
that's partly why I want to do it...
Matthias Granberry
@mgranberry
Oct 07 2015 22:12
I'd believe patents would be a problem
Ben West
@bewest
Oct 07 2015 22:12
all the stuff they want is all part of the natural process with open source communities
FDA doesn't look at patents
Matthias Granberry
@mgranberry
Oct 07 2015 22:13
of course not, but medtronic does
Ben West
@bewest
Oct 07 2015 22:13
at this point I would have a field day with medtronic suing me
Matthias Granberry
@mgranberry
Oct 07 2015 22:15
So what's the hold-up? :)
Ben West
@bewest
Oct 07 2015 22:15
oh, right, well I don't know how, would be a big part
but, I accept PRs
sometimes, I guess
Matthias Granberry
@mgranberry
Oct 07 2015 22:17
it's a shame I don't know any mechanical engineers. A backlash-free infusion pump doesn't sound like all that big a problem. I hated the medtronic screw style delivery, but it's pretty simple to do.
Ben West
@bewest
Oct 07 2015 22:18
we might be near a quorum of people for the right skill sets
Matthias Granberry
@mgranberry
Oct 07 2015 22:19
a clock-style double-latched gear driving a linear actuator would probably work.
my biggest concern with all of this is that it's tricky convincing manufacturers or distributors to work with you unless you have a war chest big enough to convince them that you're not just wasting their time.
Ben West
@bewest
Oct 07 2015 22:29

ever hear of thucydides?

Right, as the world goes, is only in question between equals in power, while the strong do what they can and the weak suffer what they must.

right now with regulatory/political issues there's a huge gray zone. it turns out that gray zone benefits the folks with the large war chests