With regard to the automation of instruments, Photosurveys (Pty) Ltd., of 
Johannesburg, have implemented an on—line mini—computer/stereoplotter 
system by interfacing a Texas Instruments 980A, 16K processor to a Wild 
A10. Stepping motors are employed to drive the plotter spindles under 
software control. The system is used for cross—section and D.T.M. measure— 
ments, where either the horizontal alignment or the D.T.M. grid is computed 
on-line and the computer ‘drives’ the measuring mark to the correct posi— 
tion, all the while transforming stereoplotter co—ordinates to terrain data in 
real time. 
A further success in automation has been achieved by the Aircraft Operating 
Co. (Pty) Ltd., of Johannesburg, who have introduced an on—line computer/ 
orthophoto projector system. They have interfaced a CDC 1700 computer 
to a Geigas Zeiss GZ1. Digital profile data is captured via three Wild A8's, 
which are on—line to the same computer, and this data is passed via locally— 
built hardware to an analogue motor which drives the Z spindle of the GZ1. 
At the same time, the Y motion of the GZ1 is monitored and in this way the 
profiles are faithfully followed. 
COMMISSION Ill. MATHEMATICAL ANALYSIS OF DATA 
RAPPORTEUR: Dr. H.S. Williams 
The period since the 1972 Congress has seen few new developments in the 
application of aerial triangulation in South Africa. There has been a marked 
tendency, however, to move toward analytical trianguiation systems. One 
commercial company has acquired the computer software of a system deve— 
loped overseas. Two systems have been developed locally, one based on the 
Schut block adjustment method using strips. Also, an attempt has been 
made to apply order systems of equations in block adjustment, but, so far, 
this has not been successful. 
As in the case in a number of other countries, interest in the simultaneous 
adjustment of large blocks of photographs appears to be waning. The high 
cost of developing programme systems for these, the solution of the many 
data editing problems, and so makes the practice of running such systems 
something of a luxury. For topographical mapping purposes, certainly, the 
adjustment of sub—blocks which requires a modicum of additional ground 
control is more than adequate. In general, blocks of photographs en— 
countered in practice usually do not exceed 200 — 300 photographs. 
Research has been carried out using three different test areas and five 
different comparator classes which has shown that a metrical accuracy as low 
as 11 microns per plate co—ordinate is adequate for obtaining currently at— 
tainable absolut? accuracies in analytical aerial triangulation. Most precision 
and universal plotters are capable of achieving standard deviations in meas— 
ured plate co—ordinates of this order. Experiments have shown that the 
rigorous Bundle method of block adjustment does not appear to produce ab— 
solute accuracies which are better than those obtainable using the Amer pro— 
cedure when applied to the adjustment of medium sized (150 photographs) 
and smaller blocks. This fact is of some considerable importance to small 
companies that only have access to limited electronic computer power. 
One member of the Society attended the ISP (Commission III) Symposium 
which was held in Stuttgart, West Germany, during September 1974. Dr. H. 
S. Williams presented an invited paper entitled "On the influence of a mini— 
mum metrical precision threshold on absolute accuracy in analytical aerial 
triangulation”.