The current answers are assuming you mean from 2018 human cities. Since it will probably take some time for our ultimate demise, some technological advances could take place that would allow a species 30 million years later to detect human presence.
It is lucky that a resourceful and vain group of scientists thought of this notion in 21XX. They developed the technology for nanobots!
Each nanobot, as part of its self replicating "DNA", contains the information from the entirety of human written works. The nanobots, while ultimately being of the 'gray goo' type, reproduce very slowly. So they themselves have been migrated by geological events. They are also fragile, so extreme heat, cold, pressure render them unable to reproduce. They have thrived though in places containing unique compounds and higher than normal carbon content (i.e. cities and garbage dumps!). Of course 30 million years is a very long time, so the nanobots also reuse the corpses of non-functioning nanobots. What is left 30 million years later, after millennia of nanobot reproduction and mutation, is areas of the Earth containing pockets of nanobots, each slowly reproducing and containing encoded information from a species long gone.
ADDED BONUS PLOT
The nanobots are quite small, and basically indiscernible from "oddly colored dirt". The information encoded in them is basically irretrievable by all but the most sophisticated of beings. The nanobots, while small, do have primitive locomotion (so they can cluster near high resource locations). They find carbon-based life forms especially delicious, particularly the large spongy masses known as "brains". They are slow to reproduce (decades), so an organism may live its entire normal life with a nanobot embedded, and not notice too much.
There is an unusual effect when a nanobot or a few hundred, invade a brain. The host organism's neural activity is altered by the molecular structure of the nanobot (which contains our encoded information). It manifests itself first as hallucinations, or wild imaginative thoughts. After much training, and the nanobots thoroughly embed themselves within the host, the host is able to enter a trance-like state where vast new worlds of information are available. This leads to the host species having cultural or technological advances that leads to...
You may be thinking that nanobots that just happen to cause changes in brain functions is a bit far-fetched and there would be a pretty slim chance this would actually happen. Unless they were designed to do that very thing!
Neural augmentation was researched for many decades to develop the technology.
It was first introduced by the clandestine, shadowy, military-industrial-complex to help create super soldiers. Of course the technology leaked into normal life. In 21XX, everyone (except poor people of course) had their memory improved by "neuraugs". They were easy to install, just take a couple of pills, and the nanobots were absorbed into the bloodstream where they were then transported to the brain.
The cutting edge research was also geared at giving new capabilities, such as:
The ability to use our existing vocal chords to more densely encode
information, and decode this information with our existing auditory
system. The neuraug soldier could communicate entire battle plans in
The ability to filter visual signals to improve vision at night or in
high brightness, or in dusty environments.
The ability to simulate sequences of physical events in a highly-parallel manner, allowing for increased performance in hand-to-hand combat.
So 30 million years later, why would the newsapiens be interested? At first, maybe they aren't, or they misunderstand what the nanobots are. At first, they appear to make people sick, and are classified similar to viruses (neither dead or alive). However, once one segment of the newsapien population discovered how to harness some of the neural augmentation capabilities, an arms race unfolds. The newsapiens don't know how to create the neuraugs, and they cannot be grown in a lab rapidly enough to satisfy the need. So they are harvested from the few sites on the Earth where they are abundant.
Since the newsapiens had a different evolutionary path from humans, their cellular makeup is similar enough for the neural encoded information to be decoded, but perhaps not as effectively and because the information was encoded from the perspective of a human host, some of the sensory signal manipulations may have novel effects that weren't originally intended. For instance, if the newsapiens have bioluminescence, or electroplaques, or echolocation, these could all be enhanced in humorous, sinister, or benign ways. (for instance, perhaps the only effect on the newsapiens is to cause a particularly attractive pattern in their bioluminescense output. So the only reason newsapiens are interested is to increase their chances at mating?)
Also, there could be mutations of the neuraugs over 30 million years, such that the neuraugs themselves have developed in to "species", each with its advantages and disadvantages.