Mr Smith BSc(Hons) GCInstCES August 2014
Membership no. xxxxxxxx
Experience Report
Site: xxxxxxx , Manchester, xxxxxx
Date: September 2013 - January 2014
Client: The Manchester Clinic
Survey Purpose: To produce accurate and up to date plans of The Manchester Clinic for space planning, reconfiguration and refurbishment to client specific specifications.
Methods and Equipment used: 3D Laser Scanning, Total station control and tacheometry, vertical control levelling, floor plan measurements with tape, distancemeter and tablet compatible software.
Personal Role: Project manager / lead surveyor
TG Surveys were commissioned to undertake an urgently required measured building survey of xxxxxxxxxxxxx, Manchester in order to produce accurate, up to date floor plans, elevations and sections of the entire building occupied by the world renowned ‘The Manchester Clinic’ in 2D AutoCAD / DWG format (GECORE01A).
The first stage of the project was to complete a full terrestrial 3D laser scan survey of the external areas surrounding the subject building (GELS01G). This was completed using a Leica C10 3D Laser Scanner with an average set resolution of 10mm at 10m (GECORE02A) which was selected to aid the drawing of detailed features and be most efficient , not considered as overkill leading to undue time spent on site and excessive data mass to handle. The 68 individually recorded ‘pointclouds’ were instrumentally controlled by a total station observed control network, using a minimum of two coordinated HDS or Black and White scan targets which were acquired from each setup location and by way of ‘cloud to cloud’ constraints also used to improve results within the chosen registering software, Leica’s ‘Cyclone’.
Permanent survey control stations were established on parallel and adjacent footpaths surrounding the building, forming the survey control network which was observed using a 1” Leica TS15 total station with 1mm +1.5ppm EDM precision. In order to ensure the desired tolerances were met, a minimum of three rounds of angles were recorded from each survey control station with more recorded where required to increase the level of measurement redundancy which would be key in calculating the control network via a traverse with level computation least squares adjustment within the chosen survey software package, ‘SCC’ (GESL03D). Each scan target was also recorded using this process as well as being observed from more than one total station set-up location where possible in order to increase the reliability of each coordinated target. Conclusions drawn from the survey report indicated that a good survey control network had been established with excellent misclosure results and that scan targets that had been observed from more than one total station set-up location were well within the desired tolerances.
These positive results were further highlighted whilst registering the 68 individual 'pointclouds'. The scan targets that were common between 'pointclouds' indicated a good registration with minimal errors which was further improved with the introduction of cloud to cloud constraints. By introducing these cloud to cloud algorithms however, the error values and calculated accuracy of the coordinated scan targets worsened which was to be expected as the cloud meshes have more common points to form a tighter bond between two individual 'pointclouds'. The cloud to cloud interim results were inspected for quality of fit and once confident that the cloud meshes were good, the weighting of the coordinated scan targets was reduced to half that of the cloud meshes. The final registration error values were well within tolerance with the average deviation of a scan target at 5mm, the largest value being 8mm, with all cloud meshes 2mm or less.
The registered 'pointcloud' was further interrogated by taking sample cross sections at various heights across the site and including surrounding buildings. By viewing the cross sections in a square-on view the registered 'pointcloud' could be seen to be tightly registered with no ‘double line’ of scan data visible which would have indicated an error in the alignment of certain 'pointclouds'.
To make the registered 'pointcloud' more manageable and useable within the chosen drafting software, Cloudworx for AutoCAD, the registered modelspace was unified with a reduction of points to one point every 5mm, to become one single 'pointcloud' rather than 68 individual 'pointclouds'. Once unified the 'pointcloud' was uploaded onto the companies designated laser scan server as well as onto a portable hard drive used for back up archiving under a unique ‘project number’ (GECORE05A). The drawing of the building footprints were then commenced before returning to site. These footprints were imported into MBS Floor Plan software to be used as base data on compatible tablet PC’s whilst on site measuring internal floor plan information. All structural openings were positioned on the footprints to ensure sufficient control points could be used for measuring internally from.
Using the unified scan data and previously plotted building footprints, the drawing of the 2D detailed elevations began. One member of the survey team was designated to this task back in the office using AutoCAD coupled with Cloudworx whilst the remaining four team members commenced with the on-site floor plan and internal measurement recording.
The floor plans of the nine storey building were recorded using a Bluetooth laser distancemeter, handheld tape measure and MBS Floor Plan compatible tablet PC (GECORE06I). Two teams of two consisting of a senior surveyor and a junior surveyor began recording the floor plan information. Generally the senior surveyor was the tablet PC operator whilst overseeing the junior surveyor recording the measurements and sending them back to the tablet via Bluetooth. This had to be carefully managed so that the two teams did not crossover and measure the same areas, although a certain amount of crossover was key to check the alignment of the two sets of floor plan data. By using the imported building footprint base data, each room / area could be accurately positioned and controlled by measuring a path externally to internally from the centre of the structural openings. These room positions were continually cross checked on site with other surrounding, independently plotted rooms and corridors by way of checking wall widths as well as the alignment of common features such as window and door positions. Another on site checking method used was the inbuilt misclosure results that are automatically displayed when closing rooms. This is calculated by the floor plan software and gives a very comprehensive check at the end of each room giving the surveyor confidence that the measurements recorded are correct. Any misclosure errors below 25mm were generally accepted and left to the software to automatically adjust the misclosure error. Errors exceeding this misclosure value were investigated before moving on with additional measurements being recorded including additional cross brace and long overall measurements.
As well as measuring the internal floor plans with the use of tablet PC’s, total station closed-traversing and detailing techniques were also used to control the main corridors running through each floor. Control stations were established through open windows at either end of each floor from externally established known control stations within the primary control network to enhance the overall accuracy of the floor plans. These internal control stations where required to be semi-permanent (for the life of the project) therefore discrete adhesive 10x10mm ‘retro’ reflective targets were used which wouldn’t be disturbed by the cleaning staff. The majority of this work had to be prearranged and often undertaken during early hours of the morning to avoid disruption to a busy working hospital.
The hard levelling of floor levels was also required throughout the building with the use of an automatic level and staff (GELS02F). The tolerances stipulated within the project specification were ±3mm between any two points used as permanent height datum’s on any floor or by more than ± 1.5mm per meter of height between floors. After completing a two-peg-test, levels were recorded from two local OS benchmarks throughout the ground floor areas whilst establishing Temporary Bench Mark’s (TBM’s) at the base of each stairwell to enable level loop closures when levelling floors above as well as the basement. Levelling the floors above (levels 1-7) proved to be a complex task as all stairwells were enclosed, not open shafts easily measured with a plumb tape or laser. In order to initially establish levels on each floor, levelling loops were completed by observing a ground floor TBM before levelling up one main staircase, establishing two TBM’s per floor (one at each main staircase) through levels 1-7 and down the other main staircase, closing the levelling loop on another ground floor TBM. Using these TBM’s, floor levels on each individual floor were recorded in a closed levelling loop before continuing to the next floor. This process was also completed for the basement level.
The internal survey tasks proved to be much more problematic due to the nature of the building use. All rooms and areas were required to be measured and positioned, regardless of their usage. Rooms such as patient recovery rooms were particularly difficult to arrange and gain access into due to the sensitivity of certain cases. This was overcome by constant daily liaising with the Ward Sister to schedule in when rooms became available to be measured. Other rooms and areas that were particularly difficult to gain access into included offices of high profile surgeons, specialists and high profile staff, sensitive archive rooms, specialist treatment rooms and surgery theatres. Access to all of these sensitive areas was often pre-arranged via liaising with department managers, secretaries etc. to be able to collect the required survey data. All site equipment, PPE and personal hygiene was required to be of the highest standard due to the buildings usage which was highlighted with proposed control measures throughout the risk assessment. Equipment had to be wiped with disinfectant wipes on a regular basis, hands had to be kept clean and alcohol gel applied routinely with PPE such as boots kept clean at all times. Survey team members were strongly advised to wash their hands before and after consuming food or drink (GECORE07A). With these issues in mind, a project specific method statement and risk assessment were compiled prior to the commencement of work on site.
Once all of the survey data had been recorded on site (all but certain non-accessible areas), the survey team members got to work to combine floor plan information and progress the site plotted MBS files to a more presentable set of 2D drawings with AutoCAD (GECORE04A). The total station observations recorded within the circulation areas of each floor were used to check and adjust the position of features where required. The hard levelled floor levels recorded on site were also calculated, adjusted where necessary and applied to the MBS files prior to exporting to AutoCAD format in order to take advantage of the automatic level calculation programme within the MBS Floor Plan software. Once all floor plan data was collated and all survey team members were happy with any discrepancies unresolved on site, the drawings were checked by overlaying each floor on top of the neighbouring floor(s) to check relative column, riser / shaft and wall alignments etc. A combination of the floor plans, elevations and laser scan data were used to create the required cross and long sections. By drawing the sections within AutoCAD using the gathered survey information, this was also used as a checking procedure to examine the alignment of walls and stairwells, thicknesses of floor / ceiling slabs and heights of various other features. All of the survey information proved to relate to one another seamlessly and so drawings were plotted to be checked for presentation of survey information. Once these final checks had been completed by management staff and any edits had been amended, provisional drawings were sent to the client with any non-accessible areas highlighted. Once these areas were made accessible by the client and staff members on site, they were measured and plotted onto the provisional drawings. Once complete, the drawings were then issued to the client in final 2D AutoCAD DWG and PDF format at 1:100 scale with the 3D laser scan data delivered in its native IMP format as well as PTS format on a USB flash drive (GELS04E).
2