Trapped by technology?

The Raymond Lemaire International Centre for Coservation at KU Leuven has run a heritage documentation module for post grad students since 2001. Year on year the quality and resolution of deliverables has progressed as technologies become cheaper and faster. This year data capture seeemed out of balance with observation and detail recording. Cartoon by Joris Snaet from 2016 celebrating the 40th year of the Centre at Leuven.


A look back Next year will see the 20th ArchDoc week at the RLICC. In preparation for this I have been asked to write a summary of the 20 editions of the initiative for publication.  I have collected my thoughts during ArchDoc 19 to come up with our reasons for gathering a team of international practitioners together to explore, teach and learn the art and science of heritage documentation. From the outset I have been determined to place metric tools in the hands of those that need them. Through the ’90s the deployment of photogrammetric drawings caused constant anxiety at the ‘sharp end’ of condition assessment and project scheduling, my job at the time was to ‘infil’ the missing bits and make the best of hand measurement to enhance the ‘ugly’ work of photogrammetrists. CAD was a boon for this but getting draughtsmen (and they usually were men) to accept the dawning of the digital was an uphill task.

‘Classic’ phtogrammetry forn the 80s and 90s. Measured drawing was relegated to filling holes in this stuff until CAD became a platform for drawing production.

Early outreach work (Stowe Summer School 1989 et seq) showed there was a thirst for knowledge and English Heritage Metric Survey Team began its outreach programme. By the turn of the Millennium the team could field a strong mix of the heigh tech and traditional for teaching at both the summer school and at schools of architecture and archaeology around the UK.

Aware the UK experience was not unique, CIPA provided an international perspective and at CIPA Potsdam 2001 paper after paper demonstrated the gap between capture and information need was (and still is) a global problem. Seeing English Heritage was able to contribute to skills development an approach was made at the conference to Mario Santana and the first joint EH/RLICC initiative was made in 2001.

In the years since the balance between metric tools and architectural drawing has been the backbone of ArchDoc, the metric technologies have advanced, almost to the point of being every day tools but the processes of understanding structure remains the same. Observation and investigation are aided by metric tools but not replaced by them.

Prof Tomas Coomans statement, in closing ArchDoc 2019, that Archdoc deliverables were missing key detail because students were ‘trapped by technology and decoration’ hit the nail on the head. Not for the first time I have struggled to see how, as surveyors, we can bridge the gap between efficient mass capture and effective architectural drawing. At the Dutch College chapel we lost the battle between detail and volume, the scans promised much and delivered too little too late. Photography saved the day but time for investigation and enquiry were lost and measured drawing only deployed as a last resort.

Student outcomes and outputs
ArchDoc deliverables require students to achieve learning in 3 subject areas:

  • Metric response to value and significance: what to measure?
  • Data capture: how to measure?
  • Data interpretation/ differentiation: what and how to present as a record?

The devil in the detail
I believe the single most difficult concept to grasp is the need for appropriate, selective observation, measurement and delineation of detail. It is the response to detail that determines the quality of the outcome. Metric capture technologies by themselves do not provide the examination or inquiry needed for confidence in determining historic, structural or functional context. Failures of detail capture are normal in both SFM and laser scan capture- this is their weakness, their strength is precision in 3D positioning.

For the money photography is by far the most effective tool for detailed surface capture, learning the basics of exposure can be a revelation. Spatial capture is possible via SFM and even hand held (like these deep crops) texture and condition can be assesed.

Surveying as a team is a positive experience, facilitators need to take the whole group through processes and make sure all are at the same level of understanding. Directing tasks should play to strengths: ask ‘who is happy with which task?’ Where tasks meet common obstacles the instructor needs to note this and share the issue with the whole group at an appropriate time. Breaks and group sessions (like for the drone and pano imaging) are important to help group assimilation.

TST focusses on selection, with realtime relay (by TheoLt of course) this selectivity is re-enforced.

The volume, complexity and performance of  point cloud and SFM products can operate as a negative factor in understanding as they give a false impression of completeness. It is easy to get involved in scanning and its processing at the expense of time spent on more incisive work. The anticipation of coverage can act as a stall on action.

The counsel of perfection
The kinds of precision possible with TST and laser scan can confuse. Heritage documentation is the art of the possible, in real life the tools will not be ideal, the time constants tight and budget tighter still. Output scale becomes the driver for determining performance. We apply the practical, not the theoretical to the problem.

Handling mass capture in a teaching setting
The ‘affordable’ scanner is not yet with us, the BLK 360 can produce some woeful results for its €20k price tag. MetaShape performance can raise false expectations.

Top:BLK 360 Medium density point cloud extract scannned from 15m. Middle: Canon S95 KapUA CHDK scripted exposure from a 4m painters pole. Rectified to TST trace. Bottom: Detail from measured drawing prepared at 1:50. Survey by PJ DeVos

The comparative cost of the scanner and the camera cannot be overlooked:

  • The BLK360, its support hardware and software are something like €20k with a recurring subscription overhead.
  • A camera on a pole, QGIS Georeferencer and TST rental all come in the €200 bracket.

With rectified photography around 100x cheaper than ‘low cost’ laser scanning it makes sense to spend time developing photography, TST and site sketch skills, they remain the most immediate techniques for results in a short time. If mass capture, and its control are prepared in advance the precious time we have could be focussed on the resolution of core skills which can then be applied to the ‘big data.’

Core skills:

  • Observation/ investigation/examination/analysis.
  • Sketching measured or otherwise.
  • Photography for rectification/SFM/narrative.
  • TST for photo-control: understanding scale, accuracy & precision.
  • Data Handing -digital tools CAD/ReCap/Metashape/ QGIS/ PS/Gimp.
  • Draughtsmanship: presentation at scale, understanding architectural form.
Aerial capture has never been more acessible. SFM from low cost drones is messy, even a ‘pro’ drone like the DJI Mavic has to be flown no higher than 25m abve surface for GSM of around 10mm, a threshold for 1:50 scale.

In an ideal world I would:

  • Prepare control network and mass capture data sets in advance of practical sessions.
  • Have a précis text of the historic context before we start
  • Only teach the incisive skills as ‘hands on’ let ‘how to do scanning’ wait for another day.
  • Have my lesson planning, with example data sets, rehearsed.
  • Avoid high cost tool dependency, develop a low cost tool-set for data handling, paying the subscriptions for Cyclone, FaroScene etc are not realistic for many practitioners.
  • Encourage ‘draw what you see’ thinking about architecture
  • Make sure the cost of equipment and its software dependency is known both as purchase and as rental
  • Support student confidence which can be fragile when faced with mass capture data, or complex technical procedures – by focus on the core skills.
  • Take the time to arrange the furniture to encourage participation-a comfortable workspace helps learning
  • Encourage questions, ask what students feel they need to know (beginning during and after tasks)
  • Remember to have fun, share the joy of the work whenever possible- it’s easy to forget what a miracle metric capture is!
A ‘robust’ network is required to link the parts to the whole. Having control in place before getting down to detail is preferred. Building a network backwards is not ideal (I do this all the time), working from the outside to the inside is better as the misclosure distribution of the outer loop can be over fewer, longer legs.
Are we in danger of spending more time looking at data than the building?

The reach of Archdoc is truly global, students worldwide have been empowered by their experience of the successes (and failures) of applied survey to conservation so here’s to the next 20!


About billboyheritagesurvey

Heritage worker
This entry was posted in Survey Practice. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.