I suggest that you discuss this topic with the current representatives of potential future users of your research and ask them which digital formats and physical media they are looking at as their preferred archive strategies. If those users do not have any preferences, the next best thing is to work with them to establish preferences.
If there is no clearly defined audience for your research materials, a practical answer must be based on reasonable assumptions about techniques and technologies that are expected to survive into the reasonable future.
One way of evaluating these technologies (in absence of specific requirements) is to examine standards that are being used and developed today. For example, AIIM, the Association for Image and Information Management (aiim.org), promotes PDF/A as a long-term digital format. (See aiim.org/Research-and-Public … _Standards. Unfortunately, the publications that detail these recommendation are available only to AIIM members.)
Choice of physical medium is tricky, because every physical medium is likely degrade over time.
Probably the best hedge is to archive in multiple non-proprietary formats that can be read on multiple platforms (PDF, plain text, image, etc.) and on different media (magnetic, optical, paper, as needed), using the best quality available (or affordable) such as acid-free paper and optical media that use gold instead of volatile dyes, and to avoid relying on cloud-based approaches that might or might not be in business for the long haul, as well as complicated hardware such as NAS devices, the thinking being that simpler is better (though these could of course be included as an alternate medium). In addition, multiple copies of the archived materials would be made, and stored in multiple locations that would be protected from calamity.
Before archiving, it will be useful to validate that the files are legible and not corrupted, and the copies of the archives would also be verified. And the archive would need to be updated over time as it changes (as files are edited, added, deleted, and reorganized), and perhaps even recreated if recommendations about formats and media change significantly. An index of some sort to aid in identifying what information is where might also be useful. File names should be cross-platform compatible as well.
All this could be quite an undertaking that might make backups and synching look trivial…practicality and cost would pf course influence how much of this work is done.
Finally, if your research is significant enough to outlast the the digital formats and media that you select, I think that you can reasonably assume that the future users will take action to preserve it by migrating it to contemporary formats and media that are appropriate for longer-term preservation. This point is often overlooked and instead we tend to assume that the techniques and technologies that we use today for archiving are their absolute final implementation. In the final analysis, we can only do what we believe is reasonable, and leave it to our descendants to continue (or not) to manage the archives we leave behind.
Some food for thought:
On a side note: I recall some years back reading about the dilemma faced by the designers of the U.S.'s proposed Yucca Mountain nuclear waste facility regarding preserving important information about the waste that would have been stored for thousands of years. They identified an approach that IIRC used content etched in native formats (i.e., text and graphics) into dense metal or ceramic plates, perhaps microscopically in order to achieve density of information to reduce the physical bulk required (in which case the archive would have included microscopy equipment to be used to access the information and pictographic instructions). Of course, this is an extreme case, but it was an entertaining read. I’ll post back if I can find it.
