Find new tracker hardware

[smaria]

It seems the current GSM module is no longer available (SIMCOM 5218, http://simcom.ee/modules/wcdma-hspa/sim5218/ ). We either need an alternative to this module, or potentially a re-design of the hardware of the tracker.

Aims for the re-design:

• logging via GSM and SD card
• small
• good chances of same or similar hardware being available in the future

See #4 for more information on the current tracker.

[smaria]

The current SIMCOM 5218 module is still available on ebay.
Positive:

• no need to re-do software
  Negative:
• still needs re-do of board layout
• unreliable, availability is likely to reduce over time and manufacturers potentially not trustworthy

Furthermore, there are ready to buy trackers
Positive:

• no hw work
• compact design
• cheap (ca. £50)
  Negative:
• no idea how to integrate with existing tracking software

Alternatively we can combine an arduino with a shield, preferably using a supplier with a good trackrecord of updating versions, documenting etc. like adafruit
Positive:

• good chance that similar parts will still be available in the future
• very simple pcb layout and hw assembly
• easy firmware writing thanks to standard libraries for arduino
  Negative:
• firmware needs re-writing
• costly (ca. £150 per tracker)

Parts I was thinking of for the arduino endeavour:

• Arduino + SD card: arduino mkrzero https://store.arduino.cc/arduino-mkrzero
• GPS + GPS: FONA808: https://learn.adafruit.com/adafruit-fona-808-cellular-plus-gps-breakout/pinouts

[smaria]

Some notes on tracker products:

• https://www.ebay.co.uk/itm/151857299746
• scientific applications from animal tracking might already have the perfect product for us?

[takluyver]

I think someone mentioned that the original plan, before the custom boards, was to use cheap phones sealed in a container. The phones would run an app which sends their GPS position to the server.

Positive:

• Cheap (Android phones from £40 on Carphone warehouse at the moment)
• GPS, GSM & battery already integrated
• Can be programmed to send data in the format required by the existing site.
• 3.5mm headphone jack

Negative:

• Not designed for this purpose - e.g. GPS might not function well in a box at water level, phone might put itself to sleep if no-one is interacting with it, or waste battery lighting up the screen.
• Developing and installing an Android app to send the data may be a much longer task than scriptable solutions.

[Nanoseb]

@takluyver This solution is explained in the 2015 proceedings (here). From what I understood it was discarded because an android phone has a quite low refresh rate on the gps (3s) and the positioning is not as good as on a dedicated device.

[chinchifou]

You can try this Github: https://github.com/Asmodeeus/SWARMAPP It is an Android application designed to work with the current Ruby server. As long as it basically sends data to the server, I think it meets your needs. Regarding performance issue : the main problem we encountered (compared to the SIMCOM) was more frequent loss of signal Code from: Thibault VIRAVAU [email protected] We can try to contact him to check if you can use the code. But that should be fine as it has also been developed during school projects. Le sam. 23 juin 2018 à 20:37, Thomas Kluyver <[email protected]> a écrit :

…

I think someone mentioned that the original plan, before the custom boards, was to use cheap phones sealed in a container. The phones would run an app which sends their GPS position to the server. Positive: - Cheap (Android phones from £40 <https://www.carphonewarehouse.com/alcatel/pixi-4-4.html#!colour=black&capacity=4GB&dealType=pg&tabID=specificationData> on Carphone warehouse at the moment) - GPS, GSM & battery already integrated - Can be programmed to send data in the format required by the existing site. - 3.5mm headphone jack Negative: - Not designed for this purpose - e.g. GPS might not function well in a box at water level, phone might put itself to sleep if no-one is interacting with it, or waste battery lighting up the screen. - Developing and installing an Android app to send the data may be a much longer task than scriptable solutions. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#16 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AGXZUKnUtc3-dt1XFcBAaYGj3QPWo6kuks5t_orXgaJpZM4U0x8Q> .

[takluyver]

Thanks; pity it didn't work so well.

[chinchifou]

Of course, it also depends mainly on the type of smartphone. Some low-end smartphones did not work as they tried to force the GPS signal to stay on the coast and not on the water (due to "software optimization") But even high-end smartphones did not perform as well as SIMCOM at that time (2015) Le sam. 23 juin 2018 à 20:53, Thomas Kluyver <[email protected]> a écrit :

…

Thanks; pity it didn't work so well. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#16 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AGXZUFc7Yd7RO9flg0NXgn7ejCjAH9ARks5t_o6xgaJpZM4U0x8Q> .

[smaria]

A point that pushes me further towards the arduino idea: It has nice scope for the future, e.g. we can add an interface that teams could use to get some information back for collaborative work. Definitely not something to focus on for now, but a nice thought.

I am ordering some bits this weekend to see if I can get the basic wiring worked out so all modules can talk to the arduino at the same time. Maybe I can manage to send a text with my mobile sim card.

We will also have a look at the spare parts arriving this week and see how many trackers we can actually make.

[smaria]

Part list for order (I will order from a UK reseller, but all parts are relative to the original adafruit shop):

• Adafruit FONA 808 MiniGSM + GPS https://www.adafruit.com/product/2542
  There is a more expensive 3G version of this, but the recommendation for beginners is to stick with the FONA 80* series, and it also seems to work less well with a passive GPS antenna. The module requires 2G sim cards, which may be a bit difficult to get, in which case the 3G module would be more attractive.

• GPS antenna https://www.adafruit.com/product/2461

• GSM antenna https://www.adafruit.com/product/1991

• microSD card breakout board https://www.adafruit.com/product/254

• an arduino pro mini, not sure yet if 3.3V/8MHz or 5V/16Mhz

Expected problems (I am sure there will also be many unexpected problems...):

• The pin requirements of the two modules are not compatible
• Too much memory on the arduino is required

[Nanoseb]

The microSD board takes digital pins and the GPS/GSM one uses UART, so both should be ok for an arduino at the same time.
Both module you linked work with 3.3 and 5V, so the arduino decision can be made regardless of that. However, current tackers use 3.7V lipo batteries, it is probably easier if we use the same batteries this time (hence using 3.3V arduino). As a nice side note, the FONA module has charging circuitry, so the battery can be charged thanks to the usb b port of the board.

Something a bit more worrying is that the 3G FONA version can't really do http request yet, from adafruit documentation:

We are adapting our FONA library to support the 3G chipset and right now we have SMS, calling, and basic functionality working but it will be a while until we get full GPRS TCP/IP and HTTP support. Also, the GPS is not as fast and low-power as the one on the FONA 808

So if we can find 2G sim card that will make things easier, but finding these sim card will probably be harder with time, so not a good long term plan decision.

The FONA module uses Mini SIM (not micro and not full size). Not sure about the previous version, but I guess it was full size one. So something to bear in mind when getting the sim cards.

[takluyver]

"Mini SIM" is what we all considered standard until micro & nano SIMs started appearing more recently. I think it's still easy to get those.

The original "Full size" SIM is the same size as a credit card, and I don't think I've seen anything that uses it.

https://en.wikipedia.org/wiki/Subscriber_identity_module#Formats

[smaria]

I've read through some giffgaff forum discussions on 2G sim cards for trackers. According to 2017 entries, there is downwards compatibility, and the issues are more with enabling the right features. I will get a giffgaff sim card to get started

[smaria]

Ok, this has been a bit of a roller coaster.

• The Adafruit GSM boards have no good availability in the UK, another very similar board seems wider available, but this makes the 'just use existing boards' look a lot less future proof
• Gerber files surfaced, yay
• More trackers arrived in UK
• The gerber files have less layers (4) than the physical trackers (6)

My conclusions from this: We will try and see if we can re-create the PCBs with the additional information of having some gerber files and a whole bunch more datasheets. Making a whole new set of trackers for this competition will cause more problems than it solves, but it is something we will have to face at some point in the future. If making new PCBs doesn't work out, and we have more boats than trackers, we will simply split the challenges that normally have all teams on the water (racing and group area scanning) into groups. Not the worst plan when attempting a group challenge for the very first time :)

[SylvainHunault]

On a side note, SIMCOM was bought by ublox : https://www.u-blox.com/en/investor-news/u-blox-acquires-simcom-cellular-module-product-line

May be the provide the same product under a different name.

[smaria]

As an existing commercial product, a friend suggested this:
https://www.findmespot.eu/gm/index.php?cid=100

A quick search brought up this from hackaday:
https://hackaday.com/2011/10/01/hacking-spot-personal-satellite-tracker-to-pass-more-information/

I haven't looked further into it, but it looks worth looking into.

Edit: looks like the highest data rate is a position transmission every 2.5 minutes, I believe this makes the product useless for WRSC tracking purposes.

[SylvainHunault]

Sorry for the delay. Apparently things went well for you.

Most commercial products have low refresh rates and therefore are useless for the WRSC tracking purposes. This is why the project targeted a custom board and website to handle from A to Z the time response and refresh rates.

Kind regards.

[southwolf]

Hey all, I just saw this project and I'm very interested.
Any update on this module choice?
I believe there are plenty of commercial products in China we can choose from to save tons of efforts. Would love to help investigate on this.

[SylvainHunault]

Hey all, I just saw this project and I'm very interested.
Any update on this module choice?
I believe there are plenty of commercial products in China we can choose from to save tons of efforts. Would love to help investigate on this.

Hey @southwolf,
Your help is welcome! I suggest to run your own research and share the results.
As I said, on my previous research (4 years ago) and those done by the previous students before me showed that there was no product on the market at a reasonable price with enough customization and a good refresh rate. Usually, refresh rate was per minute and costs were 50€/month.
Kind regards.

[tsaoyu]

@southwolf Sure, please share us any information on the potential tracker.
My wild guess is each of them wouldn't be too expensive (less than £50 or 400 RMB).

@smaria How is previous design effort goes? Does it worth another try this year to make more trackers when if the funding is sufficient?

[southwolf]

@smaria @tsaoyu There are lots of choices, such as this

Many of them use modules from Quectel, I like the compact design but not sure about features and power consumption. Will look deeper into it.

[tsaoyu]

@southwolf Cool. This module looks neat and I have found a similar shield here using the Simcom module. This could potentially leverage our effort on rewriting the firmware.