Decoding carelink

For an intro, see insulaudit.

Status

We can decode many of the opcodes that store configuration settings. We can also download all pages of history from the ReadHistoryData command, which contains the log of all actions the pump has taken on your behalf:

• glucose readings
• all bolus dosings, all usage of bolus wizard
• unabsorbed insulin
• alarms, etc...
• current settings

Decoding

We know how to decode time and as a result we can find and parse most records now.

And lining this up with the Carelink CSV exports reveals the nature of the contents for further analysis.

Future work

• simplify the protocol driver (see cl2.py)

  • fixing timing issues causing these problems:
    • fix comstatus problem - sometimes comes back missing
    • fix missing CRC - also comes back missing sometimes

• build out more decoders, collect more data

Help needed

Documentation of protocol, decoders, etc...

Data

• pcaps/
  • ./CareLink-Export-1350867079937.csv - One day of "use" manipulating lots of settings.
  • ./first_run.csv - Interogate basic settings
  • ./second_run.csv - Add more basals, quick bolus
  • ./third_run.csv - More bolus
• history/ - some experiments to get any historical data out of a pump
• rosetta-july-1-2006 - some initial experiments to perform some activities on the pump and observe changes to the memory dump of the pump
• basal-hist-2006 - a kind of baseline memory dump the day after rosetta-july-1-2006 activities.
• ground-start-0 - starting to get pretty rigorous about noticing differences between memory dumps, and isolating various pump activities
• new-years-day - carefully isolate activities and changes in the memory dump. See the different stages to watch what happens to the memory as we isolate each activitiy. Stage-5 results in being able to read the date-times out of the binary logs.

Tools

hexlog2binary.sh

Convert a hexdump from cl2.py into binary data. Used something like:

function new_name ( ) { echo $1 | sed -e "s/hex/binary/g" | sed -e "s/log/data/g"; }

bewest:~/src/decoding-carelink/new-years-day/rollover-month/stage-5$ for x in hex-*; do ../../../hexlog2binary.sh  $x $(new_name $x); done

list_dates.py

List records found in binary data by looking for dates.

cl2.py

Download data from an insulin pump. cl2.py src Working on a simpler version...

Covers about a dozen opcodes, including downloading multiple "pages" of historical data from a pump.

decoder.py

Run a binary file through a rudimentary scapy model.

usblyzer_filter.sh

# usblyzer_filter.sh was used to filter the raw usblyzer csv export into
# something a bit more manageable.
$ history
  553  ./usblyzer_filter.sh  first_run.csv  
  554  ls
  555  git mv first_run.csv pcaps/first_run/
  556  ./usblyzer_filter.sh  pcaps/first_run/first_run.csv  
  557  ./usblyzer_filter.sh  pcaps/first_run/first_run.csv   > pcaps/first_run/pcap.csv
  558  git mv second_run.csv pcaps/second_run/
  559  ./usblyzer_filter.sh pcaps/second_run/second_run.csv 
  560  ./usblyzer_filter.sh pcaps/second_run/second_run.csv  > pcaps/second_run/pcap.csv
  561  git mv third_run.csv pcaps/third_run/
  562  ./usblyzer_filter.sh  pcaps/third_run/third_run.csv 
  563  ./usblyzer_filter.sh  pcaps/third_run/third_run.csv  > pcaps/third_run/pcap.csv
bewest@paragon:~/src/decoding-carelink$ 

Thanks, -contributors