5.4 Targetting the In-vessel Datum System 
To ensure that the coordinates determined during a survey are 
in the coordinate system defined by the in-vessel datum system 
the datum system targets must also be included in the survey. 
This poses a major problem. Standard  retroflective 
videogrammetry targets which are usually fitted to the vacuum 
vessel datum system are not compatible with plasma operation 
and therefore must be removed from the vessel at the end of 
every shutdown. As the requirement for a remote survey is due 
to radiation levels preventing extended manned access it is 
clear that targets would have to be fitted remotely or that 
targets compatible with machine operation would have to be 
designed. The former option was ruled out on the basis that the 
time to fit 180 targets would be excessive. The principle of 
using hole targets was tested during videogrammetry trials in 
JET's training facility, where it was found that given good 
lighting and good contrast between the hole and its 
surroundings reasonable accuracy was achievable. With 
guidance from ESIC a plasma compatible target (figure 8) was 
developed. This consisted of a circular boss with an 
accurately machined 10mm hole. The inside of the hole was 
plasma sprayed with a uniform coating of (black) chromium 
oxide to maximise the contrast which will facilitate manual 
selection of the target using ESIC's software. 
Target boss 
Block chromium 
oxide coating 
Target 
N 
Accurately machined diameters 
Ne 
  
l'igure 8. Plasma compatible target. 
5.5 Trial Videogrammetry Survey 
At the end of JET's recent shutdown for the installation of the 
MKII Divertor a videogrammetry survey was carried out. The 
object of this survey was to record the status of the in-vessel 
datum system (fitted with retro-reflective targets), to record the 
as-built status of the main in-vessel components from the 
resulting digital model and therefore to prove the feasibility of 
using videogrammetry in future remote interventions. One of 
the main difficulties with this survey was in determining 
lighting parameters which were suitable for the conflicting 
requirements of seeing reflective targets and actual 
components. However the compromise reached allowed good 
visibility of all reflective targets and acceptable visibility of 
components (figure 9) in most positions so that the position of 
untargetted components could be determined to better than 
+2mm. 
6. INTERFACING VIDEOGRAMMETRY WITH THE 
REMOTE ARTICULATED BOOM 
6.1 Development of Survey Technique 
The final stage in automating the survey process involves 
JET's remote articulated boom on which the six million pixel 
camera will be mounted. An intensive period of survey trials is 
envisaged over the coming months to determine the optimum 
survey technique in terms of the intersection geometry i.e. 
camera positions which ensure satisfactory intersections 
between shots and sufficient visibility of components. Lighting 
will be studied in detail. Accuracy will be monitored by 
comparing videogrammetry surveys with CAT surveys. The 
aim will be to achieve similar levels of accuracy (32mm) with 
the plasma-compatible targets fitted to the datum system in 
place of the previously used retroreflective targets. The most 
suitable natural features of components for selection will be 
determined. When this is determined trials in conjunction with 
the boom will be undertaken to develop teach and repeat files 
which will allow the boom to go automatically to the camera 
positions required. 
The boom will be required also to position the scale bars 
necessary for determining the scale of the survey. Volume 
point bars will also be required. These are bars fitted with 
retro-reflective targets. These are required to ensure that the 
minimum density of targets required for the calculation is 
achieved and to ensure that an accurate overlap between 
photographs is achieved. These are essential and a design of 
*bar' which facilitates ease of handling and positioning with 
the boom as well as maximising the density of targets will be 
fundamental to the success and accuracy of the survey. 
Optimum position for these bars will be determined as part of 
the survey technique development mentioned previously. Boom 
handling of the bars will have to be studied and developed 
along with a method for positioning the bars and ensuring 
stability during the course of the survey. The time required to 
perform these operations will be a major factor in determining 
the most suitable methods. 
  
Figure 9. Working with the digital model of in-vessel 
334 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996 
  
  
Downloac 
This is re: 
boom. T 
additional 
system is 
incorpora 
processin; 
network c 
This can 
which cat 
After con 
structure 
certain til 
carrying « 
contact w 
them. A 
originally 
pipes. Th 
tiles (fig 
determine 
section Ww 
survey al 
within tol 
was not r 
twelve lo 
minutes 
achieved 
achieved 
to take ac 
steps bet 
with grad 
essential 
determin 
and in th 
quantify | 
certainty. 
being pr 
develope 
electrical 
Figure 1