In the first few decades of photography, NASA's astronauts had struggled with the lingering risk of having unusable photographs due to blur or images that were simply out of focus. This had impacted some early experiments where the documenting imagery was deemed unworkable because of these problems.
NASA was not alone in the worldwide desire for autofocus. Nikon engineers had been working on autofocus since the early 1980s in developing an automatic focusing system for the F3 camera. Their early efforts produced fruit with the F3AF camera released in 1983, albeit in a rather clunky way. In later modifications to the NASA Nikon F3 Small Cameras, Nikon changed the mirror box of the cameras to include the electronic contacts used in the F3AF. It was logical to think that they were planning to then fine tune the development of the DX-1 (the F3AF's viewfinder) so that it could operate in space and provide auto focus capability for the NASA F3 camera. It's not certain when, but that idea was scrapped sometime in the mid 1980s. In 1985, Nikon had released its first commercial autofocus camera aimed at the mass market, the N2020 (also marketed as the F-501) which incorporated a much better (and less clunky) method for autofocus than the F3AF. Nikon was hard at work developing a professional model of an auto focus camera, so NASA decided to hold off and wait for what would be released as the F4.
The F4 Film Camera was being developed alongside NASA's request for a 'digital' version. NASA had provided Nikon with special requirements for the 'digital-ready' body, whilst NASA engineers were busy developing a digital back and storage system. That effort would result in the F4 ESC (electronic still camera).
Nikon had learned a great deal from working alongside NASA on the F3. The camera had gone through a few iterations and evolved over time, and Nikon took what they learned and incorporated it into developing a camera system they hoped would address all the challenges NASA had been experiencing. Along with input from their regular commerical users, the F4 certainly would prove to be a leap forward in performance.
Not only did the F4 bring a reliable and accurate auto focus solution, but it also incorporated an advanced matrix metering system to help improve the measurement of light so that proper exposure could be achieved more often.
The F4, like the NASA Nikon F3 cameras, had a built in motor drive to advance the film. No longer would a commercial pro camera need a separate motor drive to be attached.
In fact, so much of what had been learned to have their cameras operate and survive in space, wound up being incorporated into the commercial release of the F4 camera, that when units were deliver to NASA, they were virtually identical to the commercial versions sold in stores; save for some special lubricants and perhaps some other internal mods.
From my research, it would appear that NASA ordered 40 F4 film camera bodies in their F4S configuration which incorporated the MB-21 high speed battery pack. They had previously ordered, I believe 14, specially modified F4 bodies for the digital (Electronic Still Camera) project.
The F4 cameras received from Nikon did not carry the NASA P/N (Part Number) and S/N (Serial Number) engravings that the F and F3 models had. Instead, decals were used to identify the units. The decal IDs would continue for all future Nikon cameras from that point onward.
The F4 retained the flexibility of the commercial F3 in being a modular system where 'parts' could be easily swapped out with different techonology for special needs and purposes.
One of these component parts that NASA frequently wound up swapping out was the basic camera back, opting for a data back instead.
The MF-23 data back for the F4 permitted the imprinting of any data specific to the film and image (i.e date, time frame number, shutter speed, aperture, exposure, etc.) directly onto the film. In addition, the back offered choice of location where to place the imprint.
Aside from the shooting data being captured, the back also included an Interval-Timer, freeze focus to automatically release shutter a preset focus position, auto exposure bracketing, exposure delay, long time exposure and stop functions. Needless to say, this was a huge leap forward in functionality and capability than just the simple ability of imprinting a date.
On the commercial retail side, the F4E camera option was made available in 1991.
The F4E consisted of a base F4 body with a MF-23 installed, however the battery pack that completed the set was the MB-23. Curiously however, NASA did not opt for the MB-23. The MB-23 offered the option to use 6 AA batteries which would have extended the operating life over the 4 AA batteries of the MB-21 which came standard as part of the F4S configuration they had initially ordered. Since the MB-21 had been around for three years already at this point, I suspect NASA perhaps decided that the cost benefit analysis did not make sense to swap out the battery packs to the MB-23 after the fact.
From what I have been able to uncover in my research, the NASA Nikon F4 first appeared on the Shuttle Discovery in January of 1992.
During its time in service, the F4S only performed as an IVA (Intra Vehicular Activity) model inside the Shuttle and was never used outside of the spacecraft on EVA (spacewalk) missions. I suspect this was likely due to the fact that most spacewalks usually used a predetermined focus with wide angle lenses anyway. The EVA missions would not obtain autofocus use until the NASA Nikon F5 took over from the NASA Nikon F3 in December of 1999.