1980s documents from Los Alamos National Laboratory and from Texas A&M University (under contract to NASA) indicate that there are plans to use "nuclear subselene tunneling machines" to melt tunnels under the Moon's surface, to make living, working, mining and transportation facilities for a lunar colony.
A 1986 Los Alamos report calls for using a fission powered, nuclear subselene to provide the heat to "melt rock and form a self-supporting, glass-lined tunnel suitable for Maglev or other high-speed transport modes." The report recommends burrowing beneath the surface because of the harsh lunar environment. (This would apply to Mars as well.) It further mentions that the tunnels would need to be hundreds, or thousands of kilometers long..." The actual subselenes would be automatic devices, remotely operated. In 1986, Los Alamos estimated each subselene could be built for about million and transported to the Moon for anywhere from 5 million to ,323 million. The price tag may seem exorbitantly high, but rest assured that there is easily that much, and more, available in the military's "black" budget for covert projects. It should be noted that the report did not specify how the subselenes and their crews would be transported to the Moon.
A 1988 Texas A&M study outlined plans for a slightly different model of lunar tunnel boring machine. The Texas A&M "Lunar Tunneler" would employ a "mechanical head to shear its way through the lunar material while creating a rigid ceramic-like lining". Essentially, this kind of machine would be a hybrid, mechanical TBM (Tunnel Boring Machine) that incorporates elements of the nuclear powered subselene. Although the machine would be nuclear powered it would have a mechanical cutter head that would bore through the lunar subsurface. Just behind the cutter head would be a "heating section" that would "melt a layer of lunar material within the excavated tunnel to a depth of only a few inches. This molten material could then be cooled to form a rigid ceramic material suitable for lining the interior of the tunnel."
The Texas A&M designers considered a couple of different muck disposal schemes. The two variants of the first called for the muck to be transferred vertically to the surface and either dumped or "sprayed" into a tailings pile. The second concept called for the use of special, tunnel dump trucks that would carry the muck out of the tunnel and dump it on the lunar surface. The designers recommend use of a SP-100 fission reactor for power, using liquid lithium heat pipes of the sort developed by the Los Alamos National Laboratory for the nuclear subterrene.
A second Texas A&M study, released in May 1988, also recommended use of a lithium cooled nuclear reactor as the power source for a lunar tunneler. In the second tunneler design, there are no mechanical tunneling components. Instead, the cone-shaped, nuclear powered tunneler melts its way through the subsurface like a subterrene. Some of the melted rock and soil is plastered against the tunnel walls to form a glass-like ceramic tunnel lining. The rest of the melted muck (called regolith) is passed out of the back of the tunneler and then carried to the surface for the disposal by the dump trucks that follow the tunneler through the tunnel.
I don't know if there are nuclear tunneling machines secretly making permanent bases and tunnels on the moon. But NASA plans certainly give cause to wonder.
And some of the images from the Clementine and Apollo missions of the numerous anomalous structures on the surface of moon, especially the near infrared images which Richard Hoagland believes show a network of structures both above and below the lunar surface, give even more cause to wonder.