Message Board Archives

Have anyone experimented with a GPS on a Raspberry Pi?

I am looking to develop some tracking software and pair it with some reliable hardware

Any experience with this?

I do not want to go with the commercial ones because of the subscription nature.

Thanks

In reply to steveo

CHEAPSKATE!!!!!

In reply to Dukes

Hello sir, thank your for your response, if you were my client I would do my best to save you money.

Please do not block the driveway with your moneybags

Where is Pelon when u need him?

In reply to steveo

Sorry I have never played around with Pi. Wanted to though.

In reply to nick2020

I have played around with a more powerful ODroid or something similar but I found out the Pi community is larger and you have more stable software and more options available. Hence I am switching to Pi.

Its just a small form-factor for a Linux computer with room to plug in useful hardware modules.

In reply to steveo

Sorry bro, was flying to meetings yesterday.

RE: GPS on a Raspberry Pi.

Very very simple, very low cost (under $100). Heck add a gyro too and you have a sophisticated device for less than $100.00

I have these items to spare in my "lab".

Do you code in Python? If so great if not, Python language is very straight forward.

Integrating both GPS and Gyroscope with PI is a few lines of code tweaking leveraging opensource packages. On a scale of 1 - 10 for a techie like you, I would say "2"

I have posted here numerous "Pi projects" over the years.... more than happy to help you bro. I even use the Pi in my solo flights (with gryo/gps and ADS-B receiver)

In reply to pelon

Kind sir, thank you for your info.

Precisely I am looking to display multiple vehicles on a map, with their locations, direction and rate of travel...the same crap done by the commercial apps, but with inhouse software.

I have never coded in Python, however my experience is wide, I have production code in C++ to Javascript, so I sneer at Python.

The thing of concern for me, is getting the data back from the GPS setup to our server, I am reading about some SIM card setup.

Any advice welcome, I will be sharing my results to all interested...the open source way

hate when people speak another language

In reply to ray

And to think as a layman I can help this guy but no he just wants advise and help from Nerds ....

In reply to steveo

This clearly illustrates the difference between men and women.

In reply to steveo

All very simple.

Question:

Do you need real time position management or "upon return"?

If real time position data is required, you either need sim based wifi or hotspot from drivers cell.

If not real time,then as the fleet returns, it can automatically upload on connection.

A Pi Zero with wifi is $15 the GPS $18 to $40.

Let me know the answer to above

In reply to nick2020

Don't talk so bad about Black.

In reply to pelon

Tracking = Real Time.. That was my understanding when I read the original request yesterday.

Why would you suggest "upon Return" as an option...? This would further complicate things, because you will have to setup Docking Station and Data Upload logic to the mix.

BTW.. I'm out of the tech game. Just throwing in my Dinosaur two cents.

In reply to Elsie

On return is easiest actually, but I will address that later. Obviously your idea of docking is not current nor cloud based...
But I'll explain that later. Wait for his needs firat

In reply to Elsie

Perhaps you don't understand that fleet management is not just about realtime data.

If you want to know where your delivery truck goes during the day, you do not need realtime. If you want to track in realtime, no problem, but just need to understand his Actual needs, rather than assume.

On the topic of docking...

Wifi and json upload to db when the vehicle returns to the office is not complicated bro. Easy and in 2019 cheap.

Nothing above is difficult.

In reply to pelon

We need both, as there are some areas in Guyana you cannot get a cell signal. Thats another one of the reasons we want to do a custom device and software, to have the software upload the stuff when back in range. We will be storing

Very insightful of u to recognize this condition.

In reply to Kay

Nerds are the most productive people on the planet...but anyway, I will tolerate this "lawman advice"

In reply to nick2020

In reply to steveo

Yes, when I saw Elsie reply, I was immediately reminded of how limited (high IQ?) individuals can be when they view a problem/seek solution.
All Elsie saw was face value. EXPERIENCE immediately told me seek the big picture, that is with fleet management, REAL TIME TRACKING is seldom needed. But that is another story.

Solution

The Hardware
Buy a PI B+ per vehicle.
Each Pi needs 5 volts so you will need a down converter from the vehicle battery system which is 12 volts.
Down converters are cheap on Amazon and I always use Drok brand with no issue.
The Pi B+ draws 1.2 amps at most in real world use, but you also need a GPS which will increase the AMP draw.

I recommend a down converter that can handle 3-5 amps which will more than accommodate your needs.

Get the best GPS you can afford. A USB GPS is cheap and freely available on Amazon. It is a no brainier plug and play to the Pi.

You will need a Case. I would recommend a project box (AMAZON) that has dimensions for the PI and the Converter. Consider a small 5v fan for the PI that will cool the Project box. Be sure to appropriate for ventilation.

Run the 12v DC from the battery to the converter. the converter to the PI. the Pi will power the GPS obviously.

Use 14 Gauge wire from the battery to the converter and place an inline fuse between the battery and the converter for obvious reasons.

Please think about placement. Where will the GPS be in relation to the pi? Dashboard gets hot so place PI under seat, GPS must see the sky.

I pause for any questions/responses

In reply to pelon


Easy there Pelon...
Why would you say that 'Real Time' tracking is seldom needed in Fleet Management ? That is not true at all.

Think about Utility Trucks making multiple Service Calls all day. Package Delivery trucks doing pickups and deliveries all day long. Have you ever tracked a package online..? All of that is Real Time.

You obviously provided good hardware suggestions for the solution. I'm sure it will work. Its very detailed.

But the man said he preferred to have both 'Real Time' and 'Upon Return' capabilities.

I interpret , to mean in range of a cell signal. There again, the closest approximation to Real Time.

I have confidence in your solution though. Your stuff is very current

In reply to Elsie
Yup
Solutions are never solely about what the client said.
Needs are the assessment starting point. Deploying a service considers the best fit, that us both efficient/economical.
Experience states that perceived needs and actual needs seldom are the same, hence the query. Steveo himself acknowledged this.

I bet he ends up using the non real time version in most vehicles, and places realtime only in those that require micromanaging. In other words, when they identify a bad driver, the can place a hotspot version in THAT vehicle, all others only need end of day as analysis.

This is just my rum soaked opinion.

In reply to pelon

Thanks bro, much practical information there. I will be ordering the hardware soon.

Software side:

Install DB of choice on PI, let Python script read GPS coordinates (and hey..ODB II data if required).
In the DB table, do a timestamp (now()) per row, with whatever data you decide to collect.

The PI can run as a MYSQL server with ease. The Python script does two things.

1. Read/Write data to the DB (vehicle id code should be attached to each record)
2. at set intervals (say 5 seconds): Look for internet connection or access point to upload to cloud DB

Use the PI to do everything - when the vehicle returns to the base location, let the PI auto-connect to the router and "do it's thing"

Of course, if you need any support on the code side, I am here to help but the above hardware/software idea is a workable, low cost solution.

Once you have the PI connecting at the base location, you can do whatever you need to do project-wise.

If you need real-time, the model is identical to above, step 2 (looking for connection to upload) does not care if it is mobile hotspot or base for upload.

NONE of this will challenge you.

In reply to pelon

Haha, I am sure you are thinking its a taxi app by now.

I havent started on it yet, but its a very interesting project for me, anything that gives me an opportunity to venture beyond traditional development is interesting.

Another thing I may look into in the future: a sensor grid integrated with drones for security/surveillance/counter-measures?? think about the possibilities there, application on large estates etc