Raw IoT Data

The Navixy Raw IoT Data API allows you to retrieve all the data sent by your devices in a raw format up to 6 months depending on your plan.

Parsed raw data — Data obtained immediately after decoding (parsing) incoming data packets, considering the protocol and specifics of the device model from which the packets were received.

On this page, you'll learn about API methods that give you access to parsed raw data, perfect for detailed analysis.

You can access various types of data that your devices send, such as:

• GPS and GSM information,

• Sensor readings,

• Status of inputs and outputs,

• Other attributes.

The Navixy Raw IoT Data API provides the following key methods:

• get_inputs: To retrieve a list of available attributes from a device to understand which ones can be requested in a file.

• read: To get a CSV file with decoded data values sent by tracking devices.

API Actions

API base path: /tracker/raw_data/

get_inputs

This API request returns the available attributes of a device. Before you dive into requesting raw data, it's crucial to understand the device's data capabilities and the specific names of its attributes.

Keep in mind that this request doesn't provide actual device data. Instead, it acts as a supplementary request to give you insights into the fields, so you know what data you can obtain subsequently.

You can get information about one device per request.

Request parameters

Request example

For example, we have a tracker with ID 123456, and we don't know which attributes are available to request in the raw data file. We will add its ID to this API request.

cURL

Response example

Response description

If there is more than one input of the same type, they are indexed (1, 2, 3...). In this case, only the input with the maximum index will be returned.

For example:

• If LLS Levels from 1 to 4 are available, lls_level_4 is returned, and it is assumed that LLS levels 1 through 3 also exist.

• If AVL IOs from 1 to 100000 are available for a device, avl_io_100000 is returned, then AVL IOs with smaller indexes also exist.

Additionally, you can get a description of all inputs from response using the endpoint.

Possible errors

• 201: Not found in the database. The tracker ID in your request may not match any trackers linked to the authorized user account. Ensure the correct tracker_id and API key of an appropriate user are used. The tracker number and user ID for which you request information will be included in the error description.

• 208: Device blocked. This error occurs if the tracker with the specified tracker_ID exists but has been blocked due to its tariff plan. For example, it may have been blocked because the user does not have enough money to cover the plan's charge.

read

Retrieves parsed raw data—values received from tracking devices by the platform. Each attribute in the exported file will be placed in a separate column.

The names and values of the inputs and status attributes returned by this request align with the names visible in when connecting to a device. You can find them in the right column, where the incoming data is decoded.

Request parameters

Instead of using the from/to parameters, you can set the interval parameter — an ISO 8601 formatted interval, for example, 2023-08-24T08:04:36.306Z/PT24H.

The response could be provided in a CSV or format file, with columns predefined in the columns parameter of the API request.

Here are the specifications for the table output in the CSV format:

• Rows are enclosed in double quotes. A double quote inside a string is output as two double quotes in a row. There are no other rules for escaping characters.

• Date and date-time values are enclosed in double quotes.

• Numbers are output without quotes.

• Values are separated by a comma character ,.

Simple columns

Simple columns are columns that contain straightforward and general information applicable to most models, such as time, GPS, and GSM data. For example: columns: ["server_time", "gps_fix_type", "lat", "lng", "speed"].

In the table below, you will find a list of all simple columns along with descriptions of what these columns mean and what they may return in the file.

In Parquet files, all enum parameters (such as gps_fix_type, mn_type, etc.) are transmitted as numbers starting from 0.

Examples:

• gps_fix_type - UNKNOWN = 0, NO_FIX = 1, HAS_FIX = 2, LAST_KNOWN_POSITION = 3.

• mn_type - UNKNOWN = 0, GSM = 1, CDMA = 2, WCDMA = 3, LTE

Complex columns

Complex columns are columns that represent various attributes of trackers, such as sensor data and states. We store them as different boxes, each with a common name. To access specific information, you first need to specify which box the data is from and then use a period to indicate the exact attribute name or input/output number. For example:"columns: ["flags.location_valid", "inputs.board_voltage", "discrete_inputs.1", "discrete_outputs.2"].

Unknown internal values will be returned as NULL in the file. The list of available internal values for a particular device should be obtained using the method described above.

The type and format for different inputs may vary, so we will provide you with the complex column name (without internal values) and its description in the table below.

If you specify a complex column without specifying an internal value, then all internal values will be returned as a JSON map (except for flags, which will be returned as an integer).

You can append complex columns with an asterisk symbol:

• inputs.*

• states.*

• discrete_inputs.*

In this case, the platform will search for all available columns within the specified data range and then request them from the database. In the resulting CSV/Parquet output, instead of the column with an asterisk, all existing columns will be shown in alphabetical order. Columns that are unavailable for the device model will not be included in the resulting file.

Example for standard searching by message time only in CSV

Suppose we need to request the following information about device 123456 for the period from October 10 to October 11, 2024: data on latitude and longitude, speed, fuel level from wireless fuel level sensor 1, calculated mileage, and the status of all discrete inputs. We do not need the message reception time by the server for our analysis. To get a csv file our request header must contain accept: text/csv and would look like this:

cURL for CSV

Response example for case 1 in CSV

Parquet file format request example

For getting a Parquet file with the same parameters add the format: "parquet" to the request body.

In the next example, we will slightly adjust the body of the request, so you can further test the downloaded Parquet file (navixy_data.parquet). The following code block is a Bash script used for downloading a Parquet file using a Linux distribution and its terminal.

cURL for Parquet

1. Switch to the root user or use the sudo prefix before the terminal commands

2. Create an .sh file using nano - nano curl.sh

3. Paste the cURL example above into the file

4. Write the file (CTRL+O -> Enter) and exit

Your browser does not support the video tag.

Example for searching by server time additionally

This API is designed to accommodate scenarios where you retrieve information from trackers to your applications within specified time intervals. Occasionally, trackers may experience connectivity issues. During such occurrences, these trackers automatically store information in their memory buffers. Upon re-establishing a connection, devices promptly transmit their stored information to the platform.

For instance:

A tracker was connected from 10:00 to 10:30. It then loses GSM signal, storing information in its buffer from 10:30 to 12:00. At 12:00, it reconnects and begins sending packets from the buffer. These packets are timestamped with message times starting from 10:30, 10:31, and so forth. However, the server time reflects 12:00, 12:01, and so on. If your program requests data from 10:00 to 11:00 at 11:00 without utilizing the server_time_filter parameter, it will receive messages only from 10:00 to 10:30. The program might not be aware that it needs to re-request this data once all data from the buffer has been uploaded.

To address such situations, there is optional filtering using the server_time_filter parameter. This ensures that your program will get all buffered information. This approach helps prevent potential data gaps and enhances the reliability of your application.

cURL

Response example for case 2

Possible errors

• 201: Not found in the database. The tracker ID in your request may not match any trackers linked to the authorized user account. Ensure the correct tracker_id and API key of an appropriate user are used.

• 208: Device blocked. This error occurs if the tracker with the specified tracker_ID exists but has been blocked due to its tariff plan. For example, it may have been blocked because the user does not have enough money to cover the plan's charge.