So first point is that there was work post '57 on new radars leading to a system that was intended first for the Or.346 aircraft, this being as FMCW system, but trials where disappointing. Treasurey approval in Oct '57 for a proof of concept 200W search only FMCW AI set aiming fora range of 50nm.
However the backup option was to move to PD and this option was ultimately put into action at the end '62 as a FMICW system. Flights proved the concept, such that this was the basis in '62 for the new X-band AI set for the P1154B, aiming for a 60nm range against a Canberra sized target with an auxillary Q-band using the same dish for use in ECM conditions.
By '64 this program was on good shape, but when the Admiralty abandoned the P1154B in the later part of the year, the program was downgraded to a research project looking at future AI and AEW technologies.
During '65 the AFVG lent new impetus to the effort but now Elliot was leading rather than G.E.C. constraints meant the X-band only system could be fitted into a trials Canberra as the indended use of a Valient for the full set suffered from the grounding of the Valient fleet. Flight trials began in Sept '66. The porgam suffered from a rationalising act in '66 as there was no need for a UK developed AI set, and so the flying was stopped in '67.
The ADV concept of the MRCA (Tornado) was on the cards from '69, and definition of the AI set for it was during '69 as a AI set using X-band FMICW of 1.5 -2kW. In 1970 the verdict was that it was not only possible but desirable to pursue a UK based solution and by '71 the earlier Elliot set was flying again in support of the effort.
Proposals by MEASL and Ferranti in '72 both being X-and 2kW FMICW with high PRF.
So I think we can answer that, AI.24 was always the designation of the new generation AI set after AI.18, but AI.25 might (and I stress might here) be an option for scaling the system down to fit the Sea Vixen's nose as the orriginal OR.346 to P1154B effort was a 36" dish and 32" sub-reflector.