metric measurements (e.g. correlated image coordinates) 
are decisive. Also the realization of the concepts in a 
computer program with an appropriate storage scheme and a 
favourable formulation of the algorithms and not least the 
recent developments in computer technology have an impor- 
tant influence on a critical evaluation of these concepts. 
For the development of general program systems for com- 
bined adjustment, however, more comparative investigations 
of possible concepts should be made. 
4. FUTURE INVESTIGATIONS 
In performing combined point determination there are sev- 
eral aspects, on which studies are being pursued, some of 
which are mentioned here. 
It is necessary to make an analysis of the sensors and the 
obtainable data, which are available now or in the near 
future, to get more realistic information about the error 
models, the inherent systematic errors and the stochastic 
properties of the data. 
In this analysis only sensor systems should be taken into 
consideration that include sufficient redundancy for test- 
ing on errors in the observations. Undetected gross errors 
or unmodelled systematic errors in the observed orienta- 
tion parameters of a photographic or opto-electronic cam- 
era can lead to hardly detectable distortions of the 
photogrammetric block. This situation is even more danger- 
ous, when an aerotriangulation with less or without ground 
control is carried out. 
Taking into account these realistic error models, e.g. for 
GPS receivers, ISS or GPS/ISS hybrids, comprehensive simu- 
lation studies on combined block adjustments should be 
performed. The results of these studies will give a model 
of accuracies, being within reach of the particular sensor 
and therefore should be a criterion for which tasks these 
data could be applied. 
With these results and other factors a cost/benefit cal- 
culation should be carried out. Based on this analysis a 
possible implementation of the sensor should be decided 
on, because the use of navigation data from airborne 
sensors in practice will depend decisively on economic 
considerations. 
Other important investigations are concerned with the 
quality of combined point determination, which are being 
pursued in close contact with practice. 
Computational aspects should be investigated, e.g. by 
first comparing the existing concepts. Then program sys- 
tems should be realized or extended, taking into account 
new trends in computer hardware and software technology. 
238