Presented by Shane MacLaughlin, Atlas Computers Ltd Dr Paddy Prendergast, DIT Overview ‘A new specification designed to address the demanding needs of engineering design in a tightly constrained urban environment, in terms of consistency, accuracy, re-usability, and quality’ Background QBN Experience pre-2005 Inconsistent quality of surveys Large amount of rework by engineers Delivered work not fit for purpose Need for improvement Background Surveyors experience Weak specification Poor statement of requirements Falling prices leading to falling quality Negative client feedback Background QBN topographic specification Developed by Atlas Computers Ltd in conjunction with Dublin City Council QBN project office Rigorously enforced consistency based on use of SCC software Based on IG75 grid Presented at Survey Ireland 2007 conference Met stated objectives but had shortcomings Shortcomings in 2007 spec Lack of explicit QA/QC procedures Lack of illustrative documentation for the contractor Based on old IG75 grid rather than new grid Based on implicit requirements through provision of SCC feature library New specification Overcome the shortcomings of the previous specification Generalise the scope to include similar works Form a working group representing the interests of all parties and including the necessary expertise to represent those interests Working Group Working Group Peter Muller, QBN Project Office Tom Curran, Dublin City Council, Survey and Mapping Department Shane MacLaughlin (Chair) and Áine Martin, Atlas Computers Ltd Dr Paddy Prendergast, Irish Institute of Surveyors Tom Mulreid, Apex Surveys and Society of Chartered Surveyors Vincent Molloy, Local Government Computer Services Board Ray Murphy and Stan Schoene, Murphy Surveys Ltd Gerry Healy, RPS Consulting Engineers Objectives Include documented QA/QC procedures to verify the quality of the delivered product Provide consistent results in terms of content, accuracy, and output Move from IG75 to ITM while continuing to support IG75 Achievable by the survey contractors Cost effective for the client Components Introduction and executive summary Specification main body Feature library User guide Sample reports Quality Assurance What is quality? ISO 9000 – Say what you do, do what you say, and be able prove it ‘conformance to requirements’ – Phillip Crosby To create quality we have to understand our requirements. ISO 8402-1986 standard defines quality as "the totality of features and characteristics of a product or service that bears its ability to satisfy stated or implied needs.“ ‘Quality is value to some person or persons’ - Gerald Weinberg Quality is subjective and exists within the context of providing value to a given audience. ‘Staying in business’ - W Edwards Deming Quality also relates to achieving the requirements with the available resources. In the context of providing services this amounts to meeting the needs of the client in a manner that is cost effective to both client and contractor. ‘Quality is predictability’ - W Edwards Deming QA and QC in the specification Ensure all client requirements are clearly and unambiguously stated Designed such that necessary checks and procedures are included to ensure that work being delivered meets stated requirements Providing visible and meaningful quality checks within all aspects of the survey Doing this in such a way that is achievable by the contractor and cost effective to the client Recognizing the need for continuous improvement through revisions to the specification resulting from feedback from all stakeholders Recognizing the limited scope of this particular specification, and the need to revisit the requirements to extend its scope Consistency - Content Feature library Naming conventions Colours, layers, styles Annotation Significance to the DTM Output conventions to CAD, MX, etc… Field implications 1,2 and 3 point features Strings and geometry Consistency - Accuracy Manifestation Missing Node Two reference points Pseudo Node Duplicate Line Overshoot Undershoot Sliver No reference point Guptill & Morrison, 1995 Consistency - Accuracy Relative & Absolute Accuracy Recorded Position 333333.333mE, 333333.333mN RA = 0.34m RMSE = ? Building AA = 0.34m RMSE = ? Actual dimension on Ground = 4.12m Recorded dimension in Database = 3.78m True Position 333333.000mE, 333333.402mN Boundary Wall Relative Accuracy - closeness between the recorded distance between two features in a database and the true distance Absolute Accuracy - closeness of recorded position (coords) to its true position (Most Probable Value) – use GPS Consistency - Accuracy Horizontal & Vertical Accuracy - Control Absolute - OSi GPS Network - 20mm for horizontal, but vertical = 2.5 times horizontal (satellite geometry) Relative - a few mm possible with vertical (points fixed), but horizontal ~10mm (due to setting up errors) Horizontal & Vertical Accuracy - Topographic detail Absolute - width of pogo stick beside wall & trying to hold pogo vertical rather than centreline ~10cm Relative - +/- 10cm at one end and +/-10cm at other end = +/20cm Hard & soft surfaces - block wall versus overgrown hedge with fence (horiz) or paved road versus ploughed field (vert) Consistency - Accuracy Accuracy (repeatability) - daily repeatability test by University of Melbourne (Gordini et al., 2006) - differences in metres between ‘true value’ and VRS solutions Eastings Northings Ellipsoidal Heights • Single shot NRTK observations are not suitable for survey control • Must be static observations, for a set period & post processed Consistency - Output Cartography TIN ground model Sections Deliverables CAD,SCC,MX Raw data Reports Main survey report Correction, reduction and adjustment details Station location Instrument set-up Digital signing Consistency – Grid System Improvement in absolute accuracy from 0.65m to 0.05m if move from IG75 to ITM (using static GPS obs @ 20kms) 0.7 m 0.5 m ITM recommended as primary CRS by Irish Institution of Surveyors in 2004 Supply data in IG75 during intervening period for legacy systems & projects Active GPS Network = Passive GPS Network = Trig Network = IG75 (Trig Network) 0.3m 0.1 m ITM IG75 (Polynomial) IG75 (7 parameter) Consistency – Grid System Benefits of using ITM Surveys are GPS compatible Computations simpler - no transformations Significant improvement in absolute accuracy Less impact of scale-factor INSPIRE compatible Using Grid in Quest On www.osi.ie Standalone on own PC Embedded in software New standard test to prove new versions from OSi & equipment suppliers Establishing Control Horizontal Control – linear surveys GPS baselines link traverses to ITM & provide extra rigor Link to IG75 trig network not permitted (discontinued by OSi) Constraints include: Maximum distance between stations = 150m & maximum distance between GPS stations = 1500m All stations must be inter-visible to at least 2 other stations AA < ±25mm @ 95% confidence & RA < ±5mm GPS Stations Traverse Stations Establishing Control Vertical Control – linear surveys Hold 1 station fixed to GPS height in middle of network Double levelling loop between stations using digital levels AA < ±20mm & RA < ±10mm * k (where k =√ distance in kilometres) = ±3.9mm @ 150m Link to Benchmarks not permitted (discontinued by OSi) Establishing Control Survey Control Report Description of observation methodology (Horiz & Vert) Description of adjustment methodology (Horiz & Vert) QA checks (SOPs) applied in the field and during processing Results - list of deliverables as well as map products Certification by surveyor that information supplied is correct & company has Professional Indemnity Insurance Copy of all raw data in standardised formats Establishing Control Survey Control Report Schedule of results of control stations ETRF89 Station No 1 2 3 4 Lat Long ITM Easting Northing IG75 Easting Northing Height Description Malin Head Establishing Control Survey Control Report Traverse Route Diagram GPS Network diagram Establishing Control Survey Control Report Station error ellipses – quantify magnitude and orientation metrics of station accuracies Establishing Control Survey Control Report Standardised location diagrams for control stations Topography RTK not allowed for surveying topographic features Density of information 10m on strings for sectioning 10m spot level interval Redundant measurements – to check & prove accuracy requirements are achieved Report – to include info on standard operating procedures Quality Control – Sources of error Gross Errors (blunders - measurement & computation) Misreading equipment, mis-recording a correct reading, casualness, verticality of pogo, GPS observations beside vegetation & buildings, rounding errors, …etc) Eliminated by using standard operating procedures Systematic errors (observation bias) Pattern in observations - causes can be identified, size of error can be quantified - elimination by equipment calibration & P Random errors (normal observation errors) Due to range of different equipment specifications and observer’s competency - accuracy can be quantified statistically Quality Control – Check Surveys Method Higher order of accuracy Redundant measurements No sight of main contract values Use of independent contractors Visual inspection for missing detail Reporting Check Surveys Reports Summaries for control and detail Breakdown of errors and analysis Problems highlighted Auditing the digital data QA check list Analysing reports Visual inspection Topography Sections 3d surface Auditing - QA Check list Naming conventions adhered to Model content is correct No duplicate points No crossing breaklines No missing breaklines Correct data included and excluded from the TIN Accurate boundary Correct annotation etc…. Auditing – Accuracy and reports Accuracy tolerances appear to have been achieved Correct grid system has been used Check corrections used are consistent Check model is consistent with raw data Check all specified items have been delivered Check all files are digitally signed and check signatures Dealing with non-compliant work Need to enforce the specification Good for the survey industry : Prices must be based around meeting the specification, accepting substandard work defeats competitive tendering Good for the client : Quality and consistency improve quickly Re-submission is at the contractors expense The survey contractor should complete the QA process prior to submittal Non compliant work should be rejected Using the specification (Contractor) Training User guide Rationale behind specific items Using SCC to complete such items Discussion of absolute and relative accuracy Using SCC to process check surveys Analysis of failed check surveys Using the specification (Contractor) Field tools Onboard feature libraries Leica Trimble PocketDTM SCC Using the specification (Client) Training Cost expectation Recommended usage QA implications for Check your requirements Accuracy Content New features by group and category Output formats Grid additions Tools SCC Crystal reports Results to date Marked improvment in quality of delivered surveys Errors discovered by QA process Inconsistent application of scale Accuracy shortfalls Incorrect naming GPS height processing Failure to provide all necessary data Missing services Revisions to specification based on results Support for GPS control Changed accuracy statement Standardising of reports Additional reports Distribution of first release Intended audience Local authorities Surveying bodies Engineers Survey contractors 500 printed copied, 1,000 electronic copies Communications with RICS and TSA Available electronically from Dublin City Council http://www.dublincity.ie/RoadsandTraffic/QBNProject Office/Pages/TopographicSurveySpecification.aspx Enhancements moving forward Moving to open standards LandXML Broadly supported by a wide range of land survey and civil engineering packages Used internationally See http://www.landxml.org/ for further information GML, to be included as part of the upcoming UK Highways Agency specification See http://www.opengeospatial.org/standards/gml Enhancements moving forward Support for related survey activities River surveys Structures, building and bridge facades Combined engineering / GIS surveys Support for emerging and changing survey technologies Scanners & LIDAR Implications of using huge datasets Feedback following release Enhancements and revisions based on project feedback Thank you Mr. Shane MacLaughlin Dr. Paddy Prendergast Managing Director Atlas Computers Ltd 15 Moyville Lawns Taylors Lane, Rathfarnham Dublin 16 Tel: 00353 (1) 4958714 Email: shane@atlascomputers.ie Web: www.atlascomputers.ie Irish Institute of Surveyors C/O Easons & Son (4th Floor) 40-42 Lower O'Connell Street Dublin 1 Tel: 00353 (1) 8720194 Email: patrick.prendergast@dit.ie Web: www.irish-surveyors.ie