I think the HSTs were the first to have an electronic sound for AWS caution instead of an air horn, which was followed on with the 56s, the 313 series, the 455s and the first batch of 317s. I think the 58s were the first to have the bell replaced by an electronic "ping", and from then on it was standard. With the fitment of TPWS which replaced the AWS equipment with a combined AWS/TPWS box, all air horns were replaced by an electronic sound, but the bell was retained. These also have the more modern self test that tests the clear sound too, so you get the caution tone and then bell when testing, same as modern trains give a tone then ping. That wasn't original with the electronic sounds though, early Sprinters etc just give the tone when testing, no ping.
As for the driver-guard systems, bells and buzzers have both been in use for some time. Typically EMUs have had bells, going back to the days of rope operated bells on trams, DMUs have had buzzers. Beeps have typically been for Intercity stock. There have been variations though, and some stock has more than one (Turbostars and Electrostars for example have bells and buzzers as expected, but some also have beepers).
The change from having mechanically generated noises (either bells or buzzers) to having electronically generated ones (sound is still made mechanically, but generated electronically, incase we're being pedantic), is something found across the board. It's not only cheaper and more reliable, and often smaller (compared to a bell or horn), but also allows a wider range of sounds and better control of volume meaning that specific conditions can be communicated much more effectively and with different priorities (you want an advisory sound to be much quieter than a "warning you must take action now unless you want to die" sound). Although not used on the UK railways (yet AFAIK), voice can also be used. Voice tends to be a manufacturer thing though. Different sounds are afterall only as good as the operators ability to insantly recognise what that sound means, wheras voice is clearer, but noises work across language barriers.