This lecture was presented at the 3D Digital Documentation Summit held July 10-12, 2012 at the Presidio, San Francisco, CA
Monitoring Using Laser Scanning – Case Study of Watts Towers
In 2011 a survey was carried for the Watts Towers structure. Watts Towers installation consists of 17 major sculptures constructed of structural steel covered with mortar decorated with broken glass, sea shells, generic pottery and tile. It was constructed in the nineteen twenties, is both a very important cultural landmark and a cultural icon for Los Angeles. Because of its age and construction it is in urgent need of an overall assessment of current condition. It has been subject to penetration from weather, seismic action and always at the mercy of vandalism.
This had to be cost effective, reasonably simple to execute and repeatable. Survey was needed to establish baseline data for all future monitoring; to provide an accurate set of base data to enable the assessment audit of current site condition, monitor both previous and future structural movement in selected parts of the structure; form the foundation of a management plan and the tool to plan and carry out a preservation strategy for the professional team.
We will also explore in a wider context how laser scanning is being used to provide base data for the foundation and execution of a project involving an existing structure.
Church: Alright, I would like to get started back with this afternoon’s lectures. And first starting back will be Christopher Gray with monitoring using laser scanning case studies of the Watt’s Towers. Chris is a chartered surveyor and international specialist in the measurement of existing, complicated, and often historic structures. His career includes Head of Survey Procurement for English Heritage, Survey Consultant to Welsh Heritage CADW, Deputy Director of Documentation of the Getty Conservation Institute, and the list actually goes on and on, it’s a very impressive list. He is currently the U.S. Delegate to CIPA, and board member for APTI, and also note he is actually the president of the Western Chapter of APT. So, I’m going to throw in a quick plug here and thank them again for the, [?], for the great reception that they hosted last night. So, without further ado, I’ll turn it over.
Gray: Can you hear me alright? Alright, so I got the amenable task of starting at post-lunch time and just really trying to keep you all awake for the next hour. But, a few musings from lunchtime conversations: Laser scanning, it’s just a tool you know, there are other tools out there. It’s true that my career is starts back over thirty-five years, so I started off hand survey with Income Plastic, I went to Photogrammetry English Heritage, which they supported, we did non-digital photogrammetry, we then did digital photogrammetry. We’ve always used technology as a part of survey at E.H. (English Heritage) and this is what, twenty years ago. But a typical survey that I would encroach upon would be a mixture of all techniques and it would include laser scanning, it would include control, it would include hand measurement, and I think we should all think that on board. These are not laser scan projects; this is an integration of what is suitable. We use laser scanning when it’s big, when there’s safety issues, and when there’s time components. If there is stuff that you don’t need there, then we probably wouldn’t use it. The other thing is that I’m a professional surveyor in the private sector. So when we get involved in a project, we try and deliver the kind of bare minimum that you guys want, which we feel that you need to have to do your project. We can’t fleet it out; there is no money out there. But what we do do is accuracy, and we hope that this should be the start of a dialogue between professional people doing documentation and the client, to find out what the appropriate deliverable is, and how we make that work.
One other thing, I do advise you all to go to the English Heritage Website. English Heritage had just completed a laser scan specification; it’s about fifty to sixty pages long. It contains best practice of laser scanning, it was done with the University of Newcastle and English Heritage and the Royal Institution of Charted Surveyors (RICS), 140,000 people have endorsed it and they had this as their standard document. If you go to the English Heritage website and look up metric survey or laser scan survey, it’s downloadable for free and it’s a superb piece of work. It contains about fourteen case studies and it pretty much contains everything from soup to nuts, from big projects right through to objects surveying, about how you do it and why you do it, and what’s the best way to do it.
I was survey manager for Watts Towers survey when I was with GBG. I am no longer with GBG; I am now with Molinar Group in Los Angeles. This was done probably about three or four months ago when I was with them, and we worked with LACMA, Los Angeles County Museum of Art.
So what are Watts Towers? Is everyone familiar with Watts Towers? This is a very kind of seminal Los Angeles piece of architecture. The Watts Towers are a collection of seventeen interconnected structures with heights up to thirty meters. I just going to read this off, this actually came off Wiki, and it is an easier way to do it than me paraphrasing. The Towers were built by an Italian immigrant worker, Simon Rodia, in his spare time over a period of thirty three years, from 1921 to 1954. The work is an example of non-traditional vernacular architecture and American Naïve art. They are iconic in Los Angeles, and they really are, if there is anyone who lives in L.A., you find that like the PBS News program, it’s the kind of background where everyone knows about it. It’s in everyone’s psyche over there. It’s kind of a mixture, really, because it’s part sculpture and part building, so no one really knows when we do conservation of it, whether we’re dealing with a sculpture or whether we’re dealing with a building. And the truth is it’s a bit of both. The Watts Towers are comprised of armatures constructed from pipes, rods, wrapped with wire mesh and coated in mortar. The main supports are embedded with pieces of porcelain, tile, and glass. And stuff that kids used to bring to him to put on. And this has been built up, as I said, for 30-40 years. He bent much of the Towers’ framework from rebar, using railroad tracks as a vice. In 1955, he gave it away, tired of the abuse he had received, because a lot of people didn’t like it, and he never came back. He died a decade later, leaving this monument.
So the challenge’s. Since they’ve been constructed there’s been structural movement due to earthquake and a lot of the structure is failing. There’s delamination of the surface, there’s maybe movement, which has never been quantified or looked at, even though there is suspicion that maybe there has been movement and it’s dangerous. It’s a very, very difficult structure that’s within some sort of framework of steel bars within steel bars within steel bars. To actually document it is incredibly difficult, and to actually work on it is also incredibly difficult.
There have been ongoing investigations and condition assessments. I think some people in this room have actually been involved in the condition assessment, the survey, and the work on it over the last ten years. A lot of that work actually has already included, laser scanning, which I think was used, really to produce models of the site.
The monument is now in the care of LACMA, Los Angeles County Museum of Art. LACMA contacted GBG to provide a monitoring strategy to gauge movement over time. So we needed to establish a benchmark for future and past monitoring. It has to be cost effective, it has to be easily repeatable, and it has to be accurate.
So this is our strategy. We knew we had to establish a permanent site control, meaning four control monuments outside the perimeter of the site. We felt that the minimum amount of data that would reflect any movement in the site would probably be six peaks out of the seventeen that are there. We would concentrate on the top two feet of each of those peaks and do those incredibly high resolution scans. The control would be delivered as coordinate data. It will be x, y, and z, an x-axis, a y-axis, and a height. We would deliver this data as 2D data along those axis. So, we would establish the axis, we would deliver orthodata of that axis. Does this all make sense? It’s kind of an ethereal point. So that when we come to complete the survey, we can scan a small part, we can tie it into the control; we can then overlap the 2D sets of data and see instantly if there has been any movement. That’s the thing behind the strategy to do this work. Any survey that we do in the future will have to be done within a day.
We are also looking to see if there has been any movement in the floor slab. The floor slab is also quite complicated as you can see, our brief wasn’t to worry about the decoration on the floor slab, our brief was to look to see if there had been any residual movement in the floor slab, had there been any in the past, and will there be any in the future? We need to have a benchmark of data of the surface of the slab and the peaks. So you can see the peaks are pretty complicated, because the peaks are complicated, and because laser scanning will go around and do three or four scans, or however many scans it took, and produce a 3D composite model of the top of the peak. By producing an orthoview, which is a view from an exterior vantage point and looking at the 2D data, we can then work out the center point of that peak, in a way that will be very difficult to monitor something which has a 3D shape. If we just did setup with a total station from one position, you would just hit something on the outside. So we have to do the whole 3D and we had to reduce that 3D into a 2D image, or a 2D data.
So we established a local control, and this is a closed traverse of four survey points, these nails in the concrete. And we put four of these around the site. A closed traverse means that we set up a total station over each of the stations, we site from one to the other, and we get the internal angles. By adding up those internal angles you run survey routines, you know, if it’s accurate or not. It’s self defining in terms of its accuracy. The local control is what will define the accuracy of the whole monument. And it has to be accurate and it has to be something outside of the monument. And I choose establishing this control probably takes half a day, it’s a very simple thing to do. So we had four of these. And this is the data that we got. We had an x, y, and z.
There was supplementary and control data taken into the monument. Now this is not permanent. When we are outside the monument, we can bang nails into the concrete and it doesn’t matter too much, but within the monument you can’t do that kind of thing. You may be able to put something temporary there, maybe a little cross in pencil or something like that. What we did, we extended the control data inside the monument and then took local data using the scan targets. And when we go back to repeat that, we will reset up on those four points and then bring that survey data back in to the monument, but maybe in different locations. Again if any of this is not clear, please shout out and we can kind of…
Unknown: Can we ask questions?
Gray: Yeah, please do.
Unknown: [Unheard question from the audience.]
Gray: Well, we can’t really. That’s why we would try and choose control points which are as far away as possible from where people are going to go, definitely not within the monument. But, you are right, and we don’t. We can do the best we can. There is no answer to that, there is no place you can put it where there will be no people, or where it will be abstracted from any kind of movement. We would hope that the places we have chosen will negate those movements. And they do really, most of the time a big area is not the area that usually moves, and if it is, then we’re screwed.
The survey was carried out using a Leica C10 scanner. We took a lot of scans, so we actually do have the site as a 3D composite scan, unified scan. But our brief wasn’t about the scanning of the site; our brief was about the peaks and the floor slab. The rest was icing on the cake, we grabbed it while we were there. So we did multiple scans around the monument. The minimum amount of scans, again, as I said, our work would be the minimum amount of data we can get that gives the client what he wants. There is no fat on this, so we worked out before hand where we can make sure we get set up to get the floor slab, knowing it was going to be a huge amount of things in front of it, and where can we set up on good strategic locations outside the slab or within the slab to get those peaks. And we produced a point clouds like this.
I hope, I assume that you have all seen tons of scans. We could have built the twirling turning around, but I think we might, by the end of this session, by the end of this week, might be sick death of seeing 3D scans floating around. So take it as a given that we did do the amount of scans that needed to be done, they were registered together, and they were unified into a single entity. And this was our deliverable. A control in the top, down below we set a surface texture on the scan data and then delivered it as a set of contours I think in intervals, so they are instantly recognizable. So when we repeat this, we will take a sample of this and we would overlap it on to it and see, hopefully, have instant record if there is any kind movement. The thinking is, you go back to the day, you do a bit of the slab, you do a couple of the peaks, you overlap them, and if there’s movement you think, “God, there’s movement, let’s go back and do it properly.” But, the one strategic day afterwards in a year or two years time, should define if there has been any movement in the project at all. Oh, what’s up here? Sorry there has been a bit of a hiccup. Sorry about this. Can you put a hold on that a moment? Sorry about that guys.
The deliverable is where we are going to be, a registered and compiled point cloud. So we deliver the point cloud to the client, a 3D polygonal mesh of the surface, and a set of controlled information further scan data for the permanent survey stations, the peaks, and the slab. So this is the planimetric view of each of the peaks, with their coordinate values. This is the peak information, and again, this is always taken along a specific x y axis so when we go back to the site, we will repeat that x y axis and we will overlap the data. And it will still be only 2D data; we won’t need to model any 3D data. So that was delivered and the client was happy.
But then, three months later, we had a call from LACMA, that they actually had found in their archive some original survey data. The surveys were carried out by a company called FW Cook that were carried out in 1988 and were actually repeated in 1994. The data was good, it was good survey practice, but ultimately it was before the time of EDM, electronic distance measurement, a Total Station gives you EDM. It would give you an angle and it would give you a distance, from the prism. Back then you didn’t have the distance; the lasers thing wasn’t invented then. So all you would have is horizontal and vertical angles. This is the data that CW Cook had to the proposal therefore, was to reestablish our traverse control, to go down to the site and see if we can find any evidence of the control positions that they used to do the original survey; take observations onto the newly created control points and then put the angles through the survey software, and then generate new coordinate and benchmark values of the survey that had just been found. The oldest survey then will be generated in that single system in which we had already established through the survey control.
So that shows the original control and that shows the control that we found and we showed that when we went back to the site, we’ve been using a total station only, we didn’t need to use the scanner. So all we would do is just fix those three survey points which we located and just give them x, y, and z values and then use those x, y, and z values using the angle data, to work out the peak values. So, the work was carried out in half a day. The observations, we found three control points, we were able to transcribe all the values and the data was regenerated into a single coordinate system. I haven’t gotten a copy of the data that was delivered, but the movement was detected, there were things, we call in survey, gross movement, which is something over five centimeters, which means it isn’t a fault in the survey, this is actually something that happened, and we realize that on two of the peaks, they actually had movement, was actually carried conservation on them. This complied with the data that we were finding. On the other towers, we were soon to see that the movement was five millimeters, which was an acceptable level, which is what we would expect to see.
So, conclusions. We feel this is a perfect use of the laser scanner, to generate the data that we needed. It was cost effective, it is accurate, it allows a repeatability in a half day, and it also carries out best practice survey methodology, with experienced people that know how to do this kind of work which we feel is key.
Christopher Gray is a chartered surveyor and an international specialist in the measurement of existing complicated, often historic structures. His career includes, head of survey procurement for English Heritage; survey consultant to CADW (Welsh Heritage); Deputy Director of Documentation at the Getty Conservation Institute; Director of Preservation and 3D Survey at GBG; and currently Director of Business Development for the Mollenhauer Group in Los Angeles and the UK.
Chris has managed the survey of hundreds of historic buildings and include Stonehenge, Dover Castle, The Hieroglyphic stairway at Copan, the Rose Bowl in Pasadena; throughout his career he has championed appropriate uses of technology from terrestrial photogrammetry, laser scanning and currently exploring use for BIM and 3D modelling.
He is the current US delegate to CIPA; past Board member for APTI, president of WCAPT, Fellow and member of the Global Board of Directors for the Royal Institution of Chartered Surveyors.