TIMBER MEASUREMENTS SOCIETY

Transcription

1 TIMBER MEASUREMENTS SOCIETY Fall Meeting Octber LOGMETER 4000 An example f whle-lad scanning technlgy applied t Wd Measurement Christian Pacct Wdtech, Chile

2 Agenda Wdtech Lgmeter Hw it wrks 3D Presentatin Descriptin f Measurement Prcess Mathematical Mdels Sme Installatins Results and Limitatins Repeatability Precisin Autmaticity Periphery Lgs Benefits Cnclusin Wdtech

3 Wdtech Specialist in Autmatic Wd Measurement Systems since 1990 Created frm Excelsys assciatin with the largest Financial Grup in Chile - main sharehlder f ARAUCO Prducts specifically fr wd measurement n Trucks Measurement f slid and frame vlume, as well as the bimetric characteristics f the wd. Presence tday : Latin American Prduct fr the Wrld Main Partners Strategic partnership with: SKS Visin Systems (Finland) Mgai ( Brazil ) Endeavr enterprise D

4 Clients ARGENTINA BRASIL: CHILE USA ECUADOR D

5 Agenda Wdtech Lgmeter Hw it wrks 3D Presentatin Descriptin f Measurement Prcess Mathematical Mdels Sme Installatins Results and Limitatins Repeatability Precisin Autmaticity Periphery Lgs Benefits Cnclusin Wdtech

6 LOGMETER HIGH-PRECISION LASER SYSTEMS FOR THE MEASUREMENT OF MEDIUM AND LARGE MILLS WOOD LOADS Measures: Slid vlume, frame vlume, and bimetric characteristics f the wd Autmatic peratin PORTABLE LASER SYSTEM FOR THE MEASUREMENT OF SMALL MILLS AND FOREST OPERATIONS WOOD LOADS Measures: Frame vlume f wd lads Easy t install and mve

7 Lgmeter 4000 Characteristics Slid and frame vlume Variable-length Shtgun Chips Crsswise 3D laser technlgy 50,000 captures per secnd per laser Pulp and saw wd Shtgun and crsswise lads Fixed-length, variable-length, and treelength Bimetric infrmatin fr individual lgs: number f lgs, lengths, diameters, sweep, and crk Measurement f chips and bimass Advanced auditing system Integratin with the mill s ERP system Easy-t-use and mdern user interface Tree-length

8 Agenda Wdtech Lgmeter Hw it wrks 3D Presentatin Descriptin f Measurement Prcess Mathematical Mdels Sme Installatins Results and Limitatins Repeatability Precisin Autmaticity Periphery Lgs Benefits Cnclusin Wdtech

9 Lgmeter 4000

10 Measurement Prcess Stage 1: Capture Captures the truck in mvement secnds per capture. * 150,000 samples per secnd. Capacity f up t 960 trucks per day. The picture is stred autmatically fr auditing. Berneck installatin, Brasil D * Surce: Nueva Aldea mill (Arauc), Chile

11 Measurement Prcess Stage 2: Prfiles As the truck passes, prfiles are captured. Cuts every 5cm aprx. 400 cuts are generated fr a standard lad. Millimetric precisin high definitin. D

12 Measurement Prcess Stage 3: Frame vlume Gruping f the prfiles generates a 3D image. Sphisticated image recgnitin algrithm separates the wd frm the truck. Calculates the frame vlume taking int accunt mre than 1,000,000 measurements. 99% accuracy certified by DICTUC.* D *DICTUC (Dirección de Investigación Científica y Tecnlógica de la Pntifícia Universidad Católica de Chile) cmmercially develps and implements prjects f innvatin and technlgy transference in natinal and internatinal markets.

13 Measurement Prcess Stage 4: Bimetric variables Periphery lgs are mdeled. Each lg in the periphery is an individual 3D bject. This allws fr the calculatin f Lengths Diameters Curvat Stacking ure length diameter curvature stacking etc. D bimetric variables:

14 Measurement Prcess Stage 5: Slid vlume Identificatin f characteristics fr each lg in the periphery. The slid vlume is calculated with a mathematical mdel using the Heig ht afrementined variables. An individualized mdel is calibrated fr each species in each mill in rder t ensure accuracy. The mdels can be updated Width Length D peridically.

15 Measurement Prcess Stage 6: Results and auditing Results are sent t the mill s ERP (including pictures and 3D images). All f the data is accessible by Internet. The auditing system allws fr users t filter the data by certain criteria and even electrnically measure a truck (in the case f dubts, fr example). D

16 Lgmeter 4000 Lg identificatin

17

18 Lgmeter 4000 Real capture

19 Lgmeter 4000 Real capture

20 Lgmeter 4000 Real capture

21 Lgmeter 4000 Real capture

22 Lgmeter 4000 Real capture

23 Lgmeter 4000 Real capture

24 Lgmeter 4000 Real capture

25 Mathematical Mdels Frmula: Vsl V ext X1 2 X 2 3 X 3 4 X 4 5 X 5... Vext is typically : Fr a Pulp mill 10%-20% f ttal vlume Fr a Saw mill 40-60% f ttal vlume Example f a study fr US prspect: Fr these samples, 46% t 65% f lgs are n the periphery An easy figure t remember: 30 cm diameter 50% wd is n the periphery

26 Mathematical Mdels Frmula: Vsl V ext X1 2 X 2 3 X 3 4 X 4 5 X 5... Other terms: i, i are regressin cefficients X i, variables d nt depend n the lg: Lad prfile every 3cm Average width Average diameter Diameter distributin Sweep & crk Pile arrangement Amngst thers

27 Lgmeter 4000 The bjective f acquiring a Lgmeter is t have a trustwrthy and effective measurement that can be relied upn and that ensures fair deals. Gunther Hinrichsen, Nueva Aldea Mill Fiber Superintendent, Celulsa Arauc, Chile

28 Agenda Wdtech Lgmeter Hw it wrks 3D Presentatin Descriptin f Measurement Prcess Mathematical Mdels Sme Installatins Results and Limitatins Repeatability Precisin Autmaticity Periphery Lgs Benefits Cnclusin Wdtech

29 A few installatins Date f installatin: April 2007 Place f installatin: Cnpacel Mill (ex-ripasa), BRAZIL Annual vlume scaled: 3,0 millin m3

30 Date f installatin: March 2007 Place f installatin: Nueva Aldea, CHILE Annual vlume scaled: 3,8 millins m3

31 Date f installatin: January 2006 Place f installatin: Astex, CHILE Annual vlume scaled: 0.65 millin m3

32 Date f installatin: May 2008 Place f installatin: Berneck, BRAZIL Annual vlume scaled: 1 millin m3

33 Agenda Wdtech Lgmeter Hw it wrks 3D Presentatin Descriptin f Measurement Prcess Mathematical Mdels Sme Installatins Results and Limitatins Repeatability Precisin Autmaticity Periphery Lgs Benefits Cnclusin Wdtech

34 Frame vlume accuracy Hw is frame vlume accuracy verified? Validatin bundle: A structure f PVC tubes that simulates a bundle f wd. The vlume f the bundle can be mdified by adding tubes and it is easy t measure the bundle precisely by hand.* Adjusted manual measurement (AMM): Manual methd in which up t 50 height, width, and length measurements are taken (versus less than 10 with nrmal manual measurement). * A validatin bundle is included with all Wdtech slutins.

35 Frame vlume accuracy Results Cmparisn f Lgmeter vlume cmpared t adjusted manual measurement Wd Lengths 2,4m 4,0m & 6,0m* Variable Number f Trucks Maximum Difference 3,2% 2,9% - Average Difference 0,2% 0,3% 0,6% Manual errr 2,4m 4,0m & 6,0m Variable Lgmeter difference Surce: Nueva Aldea, May and June, 2007

36 Slid vlume accuracy Accrding t results in mills, the distributin f differences curve is the fllwing: Perfrmance measurements: Average f the differences 0%. Standard deviatin f the differences 3%- 95% pssess a difference less than 5.0%

37 Repeatability Objective: Validate that Lgmeter measurements are repeatable. Test Prtcl: The validatin bundle is laded n a truck. The truck is measured n times (+/- 10) by the Lgmeter The measurements (frame vlume) are cmpared t the exact knwn vlume.

38 Repeatability Validatin bundle Test PRF_RP_001 Ripasa Mill, Ripasa, Brazil 2007 Results Precisin Trials

39 Repeatability Validatin bundle Test PRF_RP_001 Ripasa Mill, Ripasa, Brazil 2007 Cnclusin: A measurement that is nt repeatable cannt be precise. When measuring the validatin bundle, the Lgmeter cnsistently measures the frame vlume accurately, with an average abslute difference f 0.45%. Lgmeter s measurements are highly repeatable and precise.

40 Mathematical Mdels Frmula: Vsl V ext X1 2 X 2 3 X 3 4 X 4 5 X 5... Vext is typically : Fr a Pulp mill 10%-20% f ttal vlume Fr a Saw mill 40-60% f ttal vlume Example f a study fr US prspect: Fr these samples, 46% t 65% f lgs are n the periphery An easy figure t remember: 30 cm diameter 50% wd is n the periphery

41 Mathematical Mdels Frmula: Vsl V ext X1 2 X 2 3 X 3 4 X 4 5 X 5... Other terms: i, i are regressin cefficients X i, variables d nt depend n the lg: Lad prfile every 3cm Average width Average diameter Diameter distributin Sweep & crk Pile arrangement Amngst thers

42 Agenda Wdtech Lgmeter Hw it wrks 3D Presentatin Descriptin f Measurement Prcess Mathematical Mdels Sme Installatins Results and Limitatins Repeatability Precisin Autmaticity Periphery Lgs Benefits Cnclusin Wdtech

43 Main Benefits Directs All Wdtech slutins ffer the fllwing direct ecnmic benefits: ACCURACY: Traditinal measurement methds tend t verestimate the quantity f wd received, thus increasing transprt and harvest csts. The Lgmeter 4000 takes mre than 1 millin measurements per truck. As it mre precise, it des nt verestimate the lads. SPEED: Reduces peratinal csts as it requires less peratrs. Als reduces truck waiting times, allwing fr mre transprt cycles. AUTOMATICITY: Des nt invlve many prcesses nr des it require specialized peratrs, thus minimizing data-entry errrs. AUDITING: Allws fr later revisin f all results as well as remte access fr suppliers and fr headquarters. This reduces the pssibility f fraud and lss.

44 Main Benefits Indirects All Wdtech slutins ffer the fllwing indirect ecnmic benefits: TRANSPARENCY: Reduces cnflicts with suppliers as the measurement prcess is bjective. NO SAMPLING: All trucks are measured. This eliminates cnversin factrs and uncertainty regarding sample representativity. SAFETY: Zer risk fr the peratr as he/she des nt need t get n tp f the lad. ECO-FRIENDLY: 100% safe fr the envirnment des nt use fuel r lubricants.

45 Crrectly quantify wd is imprtant...but far frm simple There is n perfect methd But there are advancements.

46 THANKS