How to read the Herculaneum scrolls
Currently, there is no single method or type of scanning that can allow us to read the scrolls; the only way to achieve this is to proceed with different approaches and analyses and then add the individual results with the leverage of an AI.

Professor Brent, University of Kentucky, and Professors Berard and Charon, Institut de France, have managed to make a micrometric scan of an entire scroll, its entire surface including the internal folds.
Their main objective was to reconstruct the surface, to unroll it but, following the same procedure, by doing various types of scanning it would be possible to directly detect the ink of each single letter, through the chemical traces of the various elements that composed it: carbon, iron, gum arabic.

At the beginning, you only obtain 3D images of point clouds but, fortunately, they are not arrangements in a random, chaotic way; with the help of artificial intelligence, as you unroll the result of each scan, you can reorder the individual points into rows and pages then, by overlapping the results of the various scans, by “joining, adding” the various points and traces, it should be possible to outline the shape of many letters and, with the help of cryptographers and linguists, reconstruct entire words and sentences.

What to look for:

1) Carbon
We know that Roman ink was carbon-based, with certainly different atomic or molecular characteristics from that of incinerated papyrus; the first is made from burned wood, while the same is derived from the "baking" of plant fibers.

2) Iron
It undergone transformation processes which perhaps generated characteristics which are difficult to find in papyrus. Since it is a metal, the use of X-rays or instruments capable of measuring micromagnetic variations could also help.

3) Gum arabic
Perhaps, even the last component could have left specific traces during carbonation.

4) Impurities
Some ink lots may have been contaminated or the basic components have highly specific characteristics (for example, isotopes) that could facilitate the reading of some rolls.

5) "Used" ink
Dr. Stani and Gigli, CERIC Institute, analyzed some "intact" ink found still in the inkwell; maybe, after application on the rolls, some specific characteristics are created, such as oxidation.

6) Papyrus
Perhaps, during carbonization, the papyrus produces specific substances that could give us a negative image of the points where the presence of the ink creates differences, in a similar way to what Professor Mocella, ISASI Institute, and Professor Brun of the ESRF Institute did, who detected the variations in thickness resulting from the application of the ink.

With the help of artificial intelligence, after unrolling and overlapping the various scans, we should find ourselves a series of lines of equal length, made up of various dots, stains, traces, and empty spaces.
The traces are like a code to be deciphered.
At this point, a cryptographer, always with the help of an AI, could identify the blanks to establish the length of individual words. It would also be possible to measure the size of individual characters, similar to PC writing programs, to reconstruct individual letters; for example, the “I” is the smallest letter, it takes up the least space, while the “M” should be the wider; the “O” and the “Q” are circles; the letters “A, V” also have specific shapes; the vowels are the most common; etc. etc.

An ancient Latin linguist could make the final, crucial contribution. First, he would identify the most common articles and conjunctions. Then, even with just three or four letters, he could reconstruct many words; by coordinating words following grammar and syntax, you could reconstruct others (once a word is identified as a verb, the last letters and the words immediately before and after it, unlike in English, are coordinated according to the subject and tense); the logical sense of the sentence can also help you identify words.

Conclusions

First step: Protect the rolls in a transparent silicone bath, like the one used to preserve insects.
Second: Scan the roll in 3D, using all possible methods, looking for traces of the various ink components.
Third: Unroll and overlay the various scans, distribute the 3D trace cloud into pages and lines.
Fourth: Reconstructing the shape of the letters, divide the lines with dividing spaces to identify the words.
Fifth: Follow the grammar rules to reconstruct meaningful words, the syntax to coordinate them, and the logic to give meaning to the writing.

In the end, perhaps not all the words will be identified, but at least the content will be, and this could be useful for research: an unknown Homer text should be given priority over an annual account of the collections, and future excavations could also be better directed if the division of the library's sections was understood.