After having written the data management plan or, in any case, a document describing the treatment of data, the data collection will start. These can be data generated during the research and or data from other sources (in this case check the status of the rights and if there are any limits to the re-use and which ones)
Project and data documentation:
add to the datasets all the documentation needed to understand their treatment. If possible, accompany the dataset with a readme txt file.
File formats:
please choose when possible open formats (data formats not tied to specific software) or standard formats to ensure long term accessibility
Organisation of files and folders:
Choose meaningful file names, use a logical folder structure
Keep track of changing in your dataset using a versioning system
example: add a number to file names to indicate important changes: img.v01, img.v02, img.v03
Methods and data quality:
Describe how data were collected, tools and methodology, and how data quality was assured (calibration, measurements repetition)
Quality control (data-archive.ac.uk)