Unlocking the potential of harvester on-board-computer data in the South African forestry value chain

Date
2019-12
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: The South African forest industry is in a state of change from motor-manual to fully mechanised harvesting systems. This is predominately driven by health and safety concerns related to motor-manual harvesting systems, and the need to enhance systems productivity and product quality.Through the use of technologically advanced harvesting machinery with on-board computing systems, and standardised and compatible data collection software, all mechanised processing operations are able to produce real-time (time-stamped) data related to almost every action or function of the machine. The software referred to above is the Standard for Forest Communication (StanForD) first developed by Skogforsk in 1987, as a standard for managing the information flow from the forest machines through the value chain.Although most machines in South Africa are compatible with the StanForD systems, the usefulness of the concept remains under-utilised due to limited understanding of the interface between harvester heads and the computing systems. This includes validating the integrity and accuracy of the data emanating from the system, and that is firmly embedded in quality assurance and computer calibration. The objective of this study is to propose and develop an applicable bark deduction method for Pinuspatula in the Mpumalanga Highveld region of South Africa for more precise log volume calculations. This was accomplished by modelling historical P. patulabark thickness data from the Mpumalanga Highveld region to obtain bark thickness estimates for the two methods of bark deduction to be assessed that are available on the Ponsse Opti OBC system. Three trials were run: T1 (status quo no bark deduction function), T2 (length-based [LB] bark deduction method) and T3 (diameter-class length-based [DLB] bark deduction method). The two bark deduction methods were implemented successfully, and the harvester`s under bark (UB) diameter measurements compared well with manual measured UB diameter measurements which was derived through the novel application of photogrammetry technology. Results showed that if no bark deduction method is used the harvester over-estimates stem volume by 13.7% and 14.6% for each of two respective bark deduction methods. Furthermore, by the nature of P. patula bark being extremely thick at the base of the tree stem, means this over-estimation is even greater for butt logs. The harvester over-iv estimated the log volume of the first plywood log cut by 20.8% for T1, where through the implementation of a bark deduction method the volume estimation was improved to an under-estimation of only 1.6% and 0.2% for T2 and T3 respectively. The results of this study show that by not implementing bark deduction methods the harvester`s log volume estimations are grossly over-estimated and the usefulness of theharvester`s data for value chain management is lost.
AFRIKAANSE OPSOMMING: Moderne sny-tot-lengte ontginnings masjiene is ‘n hoogs gesofistikeerd en word beheer deur tegnologies gevorderde aanboord rekenaars. Hierdie rekenaars word bedryf deur `n data format die Standaard for Forest Communication (StanForD) wat aanvanklik in 1987 ontwikkel is deur Skogforsk. Die formaat dien as die data standard vir inligtingsvloei vanaf die ontginnings masjien dwarsdeur die waardeketting tot by die saagmeul. Alhoewel die meeste ontginnings masjiene in Suid-Afrika versoenbaar is met die StanForD data formaat word die bruikbaarheid van die konsep nie ten volle benut nie. Dit is as gevolg van die gebrekkige kennis t.o.v die interaksie tussen die masjien se rekenaar en die ontginnings masjien se sny kop. Dit sluit die bekragtiging van die integriteit en akkuraatheid van die data wat uit die sisteem voortspruit met spesifieke klem op masjien kalibrasie. Die doel van hierdie studie is om ‘n toepaslike bas-dikte-verminderings-metode vir Pinus patula in die Mpumalanga Hoëveld streek van Suid-Afrika vir meer presiese blok volume berekeninge te ontwikkel. Dit was vermag deur die modellering van geskiedkundige data P. patula bas diktheid vir die Mpumulanga Hoeveld streek om die nodige geskatte waardes van bas diktheid te verkry vir die twee beskikbare metodes van bas vermindering wat op die Ponsse Opti aanboord rekenarsisteem beskikbaar was te assesseer. Drie streekproewe was uitgevoer naamlik; T1 (geen bas vermiderings metode), T2 (lengte gebaseerde bas dikte vermindering) en T3 (diameter-klas lengte gebaseerde bas dikte vermindering). Die twee bas verminderings metodes was suksesvol geïmplimenteer en die masjien se blok onder bas deursnee metings was vergelyk met die fisiese gemeete onder bas deursnee meetings wat verkry is deur die gebruik van fotogrammetrie tegnologie.Resultate het gewys dat as daar geen bas verminderings metode gebruik word nie oorskat die masjien se volume skatting met 13.68% en 14.59% vir onderskeidelik T2 en T3 oorskat word. P. patula se bas is verskriklik dik op die onderste gedeelte van die stem wat beteken dat die oorskatting nog groter is vir blokke wat onder op die stam hul oorsprong het. Die masjien het die blok volume vir die eerste veneer blok wat vanuit die stam gesny word vir T1 met 20.81% oorskat as geen bas dikte vermindering metode gebruik word nie, en met die implementiering van `n bas dikte vermindering metode is die volume skatting verbeter na ‘n onderkskatting van slegs 1.59% en 0.18% vir T2 en T3 onderskeidelik. Die resultate van die studie beklemtoon dat deur nie bas verminderings metodes te implimenteer nie word blok volumes oorskat waardeur die bruikbaarheid van die ontginnings masjien se data vir bestuur van die bosbou waardeketting verlore gaan.
Description
Thesis (MScFor)--Stellenbosch University, 2019.
Keywords
Bark thickness deduction, Harvesters (Machine) -- Calibration, Thickness measurement, Volume estimation, StanForD
Citation