I've tried looking for connections via my endocrinologists (namely the reps from those companies that they work with), groups at Purdue Lafayette (where I got my degree), and thus far haven't found any willing to give me the time of day. But truthfully, I just don't know how to go about this process because calling the R&D department of Medtronic doesn't seem like something anyone can just do.
As stated earlier, the demonstration of a hacked pump presents all sorts of possibilities to me that I would be eager to test on myself in an attempt to marry the software with the hardware.
Gil, What's the timeline for Pancreum? Any foreseeable release date for the system described on the website? Sure looks good!
My son has Medtronic's medley of devices...I've wondered for some time why these devices haven't been made to communicate with a PDA/mobile device... Do you have any insight into that? Is it just a lag in the industry?
It's curious to me that a big company like Medtronic wouldn't be racing to do the same thing that Pancreum is doing...
well.. I'm now on the road to hacking the DexCom G4.. I figured someone upstairs gave me the brains to hack at it, as to help my daughter.. so ordered some hardware to figure out the transmission protocol.
There are three goals to why I'm attempting this. A) While she's in school, I think it would be an awesome feature for the parents to get a text message on the status of their children's glucose numbers. We're now onto our 3rd week, and see things on her CGM that we didn't know (ie stress from certain teachers causing her numbers to swing.. always the same pattern with the same teachers).
B) Repeating signals. Though the range of the G4 is awesome (my daughter is upstairs in her room doing homework, we are downstairs getting supper ready), we can see what's going on. Imagine repeaters (just like wifi), throughout the home or property, while the receiver stays inside.. total freedom)
C) Smartphone implementation. The CGM tech is cool, their windows management software is pretty cool.. the receiver is major "bleh". Either build a device with Android, or port the receiver as a smartphone app.. but then again, there's money to be made and it would be hard to charge hundreds for an app...so not holding my breath..
Seems like a no-brainer to have a pump or CGM or standard glucometer be bluetooth enabled for realtime transmission of data to a cloudbased database via iPhone/Android. Push notifications could keep parents in the loop.
yeah.. you would think. The whole thing reminds me of Garmin.. they saw the writing on the wall, then tried to come up with the Nuviphone, which was Android based.. too little too late.
I mean, even the customization of alerts on the CGM..this should be pretty straightforward..
Part of me feels like an ingrate, because truth be told, the CGM is just so damn awesome.. and so what if the interface smells like 1990's tech..
Just saw your message. We're still in the engineering phase, and still negotiating with venture capital firms for more funding. So, still a few years away. There are individual products controlled by a smartphone, but not the same smartphone controlling devices made by different companies. Those companies can't agree on that YET, unfortunately.
Medtronic IS racing to develop a wearable set of devices. I've hear they've spent more than US$ 100 million, but can't get it right like we did it at Insulet.
Cheers, Gil.
Some of this stuff is a little beyond me (I'm a database guy), but I'm still hacking away...
What was throwing me off when scanning (and it was making me nuts) was in order to debug, you really shouldn't have much interference so you can isolate the signal and then start reverse engineering what they're doing.. As it turns out, I've got a bunch of wireless keyboards and mouses (I guess you'd call them Mice ?) and they're broadcasting in the 2.4ghz range.. so that totally messed me up...
ALSO.. AND THIS IS KEY.. the devices Dexcom plans on Using are ANT+ devices.. so if you ever wanted to write your own app, it would have to be for ANT+ phones (Sony/Ericson and I believe the Galaxy S3 is an ANT phone)...
What I'd like to do (and I'm giving away the keys here) is the following: A) Have a smartphone pick up the signal.. not only give the user the value, but also post via http calls (Android has libraries for this) to post to a web site (maybe run a DDNS service).. so I can check her sugar when she's in school..I thought that was a better approach, or an alternative to text messaging... Also, have the app smart enough to queue up the results when not in wifi range. The other thought I had was to basically run little devices like repeaters throughout my home and yard.. so the receiver is the "home" station and these little wireless repeaters are placed throughout my yard.. giving my daughter total freedom ...
Molten, also, I'm choosing 3g via usb modem, with wifi as back up. You hit the nail on the head... this is about freedom... how I get to choose to spend my time.
Have you captured any successful traces from the Dexcom yet? Do the current devices use ANT or have you peaked inside to see what transceiver chips they are using? It might help if we can find out more about their protocol stack rather than trying to debug the raw output from the NRF24L01.
ok.. I must have replied a while back to this.. and to be honest, I was struggling, so I kind of buried my head in the sand.. so I've been distracting myself with automated christmas lights and now just got my sainsmart solid state relays to replace my Toro sprinkler control.
If you pick up the Dexcom g4 manual, they tell you which frequencies they are broadcasting on.. and I've picked up signals from those frequencies.. My next step was to capture the raw info.. From what I understand, there's a pretty standard protocol that all these ANT+ devices adhere to (bit 1 means this, bit means that) and I'm guessing that the sensor is broadcasting a voltage measurement that the receiver is calculating to come up with a number..(some of you are probably going 'duh' of course it is, but bare with me)..
I like the idea of Github.. maybe having other people to talk about this will spark a fire under my tuchus..since I really want my daughter to have as much freedom (and freedom from worry - She's 11).
The G4 has been, hands down, the greatest thing I've ever bought.. just me a few days to respond and put things in Github.. (taxes up the wazoo and the 15th is Monday)
If you think the G4 is an ANT+ device, I would recommend using an ANT+ IC instead of the NRF24L01. Otherwise, you'd have to find or write a library to handle the ANT protocol before you get any meter data. Something like this:
That certainly would be nicer for a finished product if you can design and build the board. I haven't found any cheap modules utilizing that SoC for prototyping though. Do you know of any?
I found a quick youtube video showing someone using a pair of AP2s:
