Manufacturing Utility


This utility is designed to create per device instances factory nvs partition images for mass manufacturing purposes. These images are created from user provided configuration and values csv files. This utility only creates the manufacturing binary images and you can choose to use or Windows based flash programming utility or direct flash programming to program these images at the time of manufacturing.


This utility is dependent on the esp-idf nvs partition utility.

  • Operating System requirements:
    • Linux / MacOS / Windows (standard distributions)
  • The following packages are needed for using this utility:


Make sure the python path is set in the PATH environment variable before using this utility.

Make sure to include packages from requirement.txt in top level IDF directory.


CSV Configuration File:

This file contains the configuration of the device to be manufactured.

The data in configuration file must have the following format (REPEAT tag is optional):

name1,namespace,    <-- First entry should be of type "namespace"


First entry in this file should always be namespace entry.

Each row should have these 3 parameters: key,type,encoding separated by comma. If REPEAT tag is present, the value corresponding to this key in the Master CSV Values File will be the same for all devices.

Please refer to README of nvs_partition utility for detailed description of each parameter.

Below is a sample example of such a configuration file:



Make sure there are no spaces before and after ‘,’ in the configuration file.

Master CSV Values File:

This file contains details of the device to be manufactured. Each row in this file corresponds to a device instance.

The data in values file must have the following format:



First line in this file should always be the key names. All the keys from the configuration file should be present here in the same order. This file can have additional columns(keys) and they will act like metadata and would not be part of final binary files.

Each row should have the value of the corresponding keys, separated by comma. If key has REPEAT tag, then its corresponding value must be entered in the second line only. Keep the entry empty for this value in the next lines. Below is the description of this parameter:

Data value.

Below is a sample example of such a values file:



If ‘REPEAT’ tag is present, a new Master CSV Values File is created in the same folder as the input Master CSV File with the values inserted at each line for the key with ‘REPEAT’ tag.


Intermediate CSV files are created by this utility which are input to the nvs partition utility to generate the binary files.

The format of this intermediate csv file will be:

key,namespace, ,


An intermediate csv file will be created for each device instance.

Running the utility

The utility is using the generated CSV Configuration file and Master CSV Values file and is generating per device instance factory images.

Sample CSV Configuration file and Master CSV Values file is provided with this utility.


$ ./ [-h] [--conf CONFIG_FILE] [--values VALUES_FILE]
                [--prefix PREFIX] [--fileid FILEID] [--outdir OUTDIR]
                [--size PART_SIZE] [--version {v1,v2}]
                [--keygen {true,false}] [--encrypt {true,false}]
                [--keyfile KEYFILE]
Arguments Description
–conf CONFIG_FILE the input configuration csv file
–values VALUES_FILE the input values csv file
–prefix PREFIX the unique name as each filename prefix
–fileid FILEID the unique file identifier(any key in values file) as each filename suffix (Default: numeric value(1,2,3…))
–outdir OUTDIR the output directory to store the files created (Default: current directory)
–size PART_SIZE Size of NVS Partition in bytes (must be multiple of 4096)
–version {v1,v2} Set version. Default: v2
–keygen {true,false} Generate keys for encryption. Default: false
–encrypt {true,false} Set encryption mode. Default: false
–keyfile KEYFILE File having key for encryption (Applicable only if encryption mode is true)

You can use the below commands to run this utility with the sample files provided:

$ ./ --conf samples/sample_config.csv --values samples/sample_values_singlepage_blob.csv --prefix Fan --size 0x3000

$ ./ --conf samples/sample_config.csv --values samples/sample_values_multipage_blob.csv --prefix Fan --size 0x4000


When you use this utility to generate per device instance factory images –conf, –values, –prefix and –size arguments are mandatory.

$ ./ –conf samples/sample_config.csv –values samples/sample_values_singlepage_blob.csv –prefix Fan –size 0x3000 –outdir tmp


The –outdir directory is created if not present.


The file path given in the file type in the values file is expected to be relative to the current directory from which you are running the utility.

$ ./ –conf samples/sample_config.csv –values samples/sample_values_singlepage_blob.csv –prefix Fan –size 0x3000 –encrypt true –keygen true


keys/ directory is generated with the encryption keys filename of the form prefix-fileid-keys.bin.

You can also run the below command to use the utility to only generate encryption keys binary file ( following example ‘keys/’ directory is created in current path), which can further be used to encrypt per device instance factory images:

$ ./ --keygen true

$ ./ --keygen true --keyfile encr_keys.bin


When running utility to generate only keys, if –keyfile is given it will generate encryption keys with filename given in –keyfile argument.


When you use this utility to generate only encryption keys –keygen argument is mandatory.


The default numeric value: 1,2,3… of fileid argument, corresponds to each row having device instance values in master csv values file.


bin/ and csv/ sub-directories are created in the outdir directory specified while running this utility. The binary files generated will be stored in bin/ and the intermediate csv files generated will be stored in csv/.


Comments are supported in input config csv file only.