daunting early versions of the software. The concept of not 
having to insert or set up models again after triangulation 
represents highly worthwhile savings over APs. The export of 
orientation parameters to APs and even analogue instruments 
. Today’s DPW solutions for the commercial 
market-place are suitable, often with little customising, for 
many military needs. Indeed, military and academic users are 
the main purchasers of vendors’ modules for processing 
satellite imagery and the availability of software modules with 
mathematical models for several of the classified military 
satellites remains an important advantage of digital over 
analytical photogrammetry. 
Users’ concerns include fitting DPWs into their existing 
systems, so their demands are numerous: particular coordinate 
systems and map Projections; incorporation of GPS data 
acquired during flight; continuing use of existing block 
adjustment package; formats for image, DTM and vector data; 
exchange of orientation data with analogue and analytical 
workstations; ergonomic environment in terms of stereoscopic 
viewing, xyz control devices, etc; same software for map 
compilation as on existing analogue and analytical 
Workstations; and highly functional, heterogeneous computer 
networking of DPWs with one another, with scanners and with 
existing workstations. 
The preponderance of digital orthophotos has been followed 
after some delay by the development of standards. ASPRS 
have a committee Working on this area (Nale, 1995) and, inter 
alia, national mapping organisations like Geomatics Canada 
(Armenakis, Regan and Dow, 1995) and Ordnance Survey of 
Ireland are taking the issue very seriously indeed. Yet it is true 
to say that in 1996 standards for orthophotos are less 
Widespread and less sophisticated than those for line maps. 
This is important, because So popular has digital 
orthophotography become that many mapping companies 
promote image maps as their premier service and the focus of 
attention has progressed from feasibility to production 
problems, for example Manzer (1995). 
A debate of today concerns monoplotting. Digital 
photogrammetry opens up the chance of routinely performing 
head-up digitising on orthophotos. Indeed if one assumes that a 
DTM must have been available to make the orthophoto, then it 
follows that classical monoplotting with a raw image and a 
DTM is also possible and presumably more accurate since no 
resampling has taken place. A monoplotter can be created from 
little more than a CAD package such as MicroStation if jt has 
the facility to handle an image layer, ie. a rigorous 
photogrammetric package is not essential if an orthophoto is 
used. But it remains moot whether this approach, based on 
orthophotos which in turn are generated using DTMs of finite 
accuracy, and without stereoscopic viewing to enhance 
accuracy or interpretation, is good enough. 
The above paragraphs are enough to hint at the amazing 
diversity of purchasers of DPWs, their existing methods and 
their applications. The result has been a series of enormous 
challenges for the vendors. The DPWs have had to be 
dovetailed into existing computer networks, data flows and 
data formats. The energetic competition between vendors has 
ensured their willingness to meet stringent customer 
requirements in terms of networking, customising, workflow 
design, provision of consultancy, etc. This has resulted in poor 
vendors and some highly sophisticated users. Nevertheless, 
393 
many of these “top end” users have worked enthusiastically 
With the vendors to reach their goals in high tech partnerships. 
Competition has inevitably led to very fast development 
thought out development plans, coupled with a rather high 
dependence on beta sites. On the other hand, this charged 
environment has undoubtedly accelerated the progress of the 
DPWs. One simple example illustrates the trend: it is now 
unexceptional for a customer to place an order for DPW 
software, with a stipulation that it must run on some computer 
model on which the software has not even been tried at the 
time of order! 
Innovation diffusion is, therefore, in full swing. Aggressive 
sales and marketing by the main vendors, coupled with 
Conservatism amongst some users, especially those with heavy 
workloads in feature extraction, have ensured buoyant sales of 
APs in the period 1994-96, but it is clear that the peak has 
been passed and the decline has begun. The last page is about 
to be tumed of the AP story, a glorious chapter in the annals of 
photogrammetric instrumentation, and by 1998 it is likely that 
almost every sale of a new system will contain a DPW. 
Leberl's (1991) argument that the change from APs to DPWs 
would be faster than that from analogue plotters to APs is 
likely to prove accurate. 
4.1.4 Evolutionary or revolutionary? It is trendy these days 
to rave about paradigm shifts. Does the general adoption of 
DPWs mean that we have enjoyed such an event in 
photogrammetry? In a cogent, mature look into the future 
almost two decades ago, Ackermann (1977) contrasted 
consolidation and incremental development with major leaps 
forward occasioned by the advent of computers and electronics 
in our instrumentation. In a later overview (idem, 1992), the 
same message was repeated, from rather later in the course of 
events. At the same time, Leberl (1991, 1992a) was 
evangelistic. More recently still, Ackermann is convinced: 
“The major technical breakthrough ... is the step towards 
digital photogrammetry... the development which now has 
started is so fantastic and has unlimited potential that I am not 
at all worried about future development.. The result will 
supercede [sic] anything which we could ever have dreamt of, 
simply because of the power of digital technology" (Anon, 
1995, pp. 41 and 45). 
Ackermann (1977) and Leberl (1991) also contrasted 
revolutionary with evolutionary development. Such 
assessments are not particular to photogrammetry but are 
scattered through the advances of many sciences and 
humanities. Leberl argued that the change from analogue to 
analytical photogrammetry had been evolutionary and would 
prove to have been slower than that from analytical to digital, 
which was revolutionary despite the fact that much automation 
Was not yet operational. More recently, however, others have 
taken up the cudgels in this debate. Wong (Anon, 1996) set out 
typically forthright views, saying that we were indeed 
undergoing a revolution, that one of the stumblingblocks was 
not technical but was the lack of full understanding of the 
processes and their true costs, and, indeed, “virtual machines” 
were here since technology would enable software to convert 
any PC into a DPW, opening access to all. This idea has some 
common roots, perhaps, with Leberl's phrase “democratization 
of photogrammetry" and one wonders, too, whether the client- 
server world or the Internet will encourage in the future the 
availability of bits of digital photogrammetric software to be 
grabbed and used when they are needed. It is churlish to 
dampen such excitement, yet the truth today may be that in 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B2. Vienna 1996