Role of spatial data uncertainty in executions of precision farming operations

Jere Kaivosoja

Role of spatial data uncertainty in executions of precision farming operations

Julkaisun otsikon käännös: Paikkatiedon epävarmuuden vaikutus täsmäviljelytyötehtävien onnistumiseen

Jere Kaivosoja

Tutkimustuotos: Doctoral Thesis › Collection of Articles

Abstrakti

Tässä väitöskirjassa tarkastellaan peltoviljelytoimenpiteisiin liittyvän paikkatiedon laatua. Tyypillisesti pelto halutaan työstää joko mahdollisimman tasaisesti, tai täsmäviljelymenetelmien avulla paikkakohtaisesti suunnitellun tarpeen mukaan. Kunkin peltotoimenpiteen lopputuloksessa on kuitenkin paljon epävarmuutta. Tämä hankaloittaa erityisesti täsmäviljelymenetelmien kehittämistä sekä arviointia. Tämä väitöskirja lähestyy tätä ongelmakenttää paikkatiedon epävarmuustarkastelun näkökulmasta hyödyntäen paikkatiedon laadun standardisoituja mittausmenetelmiä. ISO 19157:2013 standardi määrittelee paikkatiedon laadun pääelementit. Tässä väitöskirjassa näitä laatuelementtejä selvitettiin tutkimalla soveltuvia peltoviljelyteknologian tapaustutkimuksia. Tällä tavoin saatiin käsitys eri laatuelementtien merkittävyydestä sekä vaikutuksesta peltoviljelyn kokonaislaatuun. Näissä teknologialähtöisissä tapaustutkimuksissa tutkittiin kaukokartoitusaineistojen hyödyntämistä täsmäviljelyssä, satelliittipaikannusjärjestelmien hyödyntämistä peltonavigoinnissa, sekä viljelijöiden käytännön työn tarkkuuksia. Kaukokartoitusaineistoina käytettiin satelliittikuvia sekä droonien avulla kerättyä tavallista näkyvän valon dataa, hyperspektridataa sekä laserkeilausta. Näitä aineistoja yhdistettiin maatilan paikkakohtaisiin tietoihin, sekä viljelijän hiljaiseen tietoon. Ajallista laatua ja aikakriittisyyttä tarkasteltiin tässä työssä toteuttamalla ulkopuolisia paikkatietoaineistoja reaaliajassa hyödyntävä ISOBUS-standardin mukaisen työkoneohjaimen prototyyppi. GNSS paikannuksen aiheuttamaa epävarmuutta tarkasteltiin luomalla realistinen GNSS-virhesimulaattori liikkuvaan ympäristöön. Tutkimuksissa havaittiin, että viljelijät ajavat työkoneiden työlinjat keskimäärin 10 % päällekkäin. Erilaisten ohjausavustimien avulla tästä voidaan vähentää noin puolet. Suurimpana epävarmuustekijänä oli täsmäviljelysäätötasojen paikkakohtainen määrittäminen. Tätä temaattista tarkkuutta selvitettiin vertaamalla eri menetelmillä samaan tarkoitukseen laskettuja täsmäviljelytyötehtäviä keskenään. Nämä työtehtävät vaihtelivat keskimäärin 22 %. Työtehtävien täydellisyys tulkittiin tässä tutkimuksessa vastaavan työtehtävän laajuutta. Ajallinen laatu oli täysin riippuvainen käyttötapauksista. Lopputulokset viittaavat siihen, että teknologian näkökulmasta peltoviljelyn ja erityisesti täsmäviljelyn kannattavuutta voitaisiin kehittää merkittävästi. Yksittäisen täsmäviljelytoimenpiteen kannattavuudeksi laskettiin noin 31€/ha, mikä oli vain n. 23 % ideaalista tilanteesta, jossa paikkatiedon laatuun ei liittyisi epävarmuutta. Käytännön tulosten mukaan työkoneen tehtävän tarkkuus oli 61 % teoreettiseen optimiin verrattuna. Lopputulosten perusteella nähdään, että teknologioissa on vielä huomattavasti kehitettävää, ennen kuin todellisia täsmäviljelysovellutuksia pystytään kehittämään ja aukottomasti arvioimaan.

Julkaisun otsikon käännös	Paikkatiedon epävarmuuden vaikutus täsmäviljelytyötehtävien onnistumiseen
Alkuperäiskieli	Englanti
Pätevyys	Tohtorintutkinto
Myöntävä instituutio	Aalto-yliopisto
Valvoja/neuvonantaja	Virrantaus, Kirsi-Kanerva, Vastuuprofessori Nikander, Jussi, Ohjaaja
Kustantaja	Aalto University
Painoksen ISBN	978-952-60-8840-2
Sähköinen ISBN	978-952-60-8841-9
Tila	Julkaistu - 2019
OKM-julkaisutyyppi	G5 Artikkeliväitöskirja

Tutkimusalat

paikkatiedon laatu
maatalous
täsmäviljely

Pääsy asiakirjaan

http://urn.fi/URN:ISBN:978-952-60-8841-9

Siteeraa tätä

@phdthesis{61385eee6f094c26b61d13c02b228f22,

title = "Role of spatial data uncertainty in executions of precision farming operations",

abstract = "The topic of this dissertation is the quality of spatial data related to arable farming work executions. There is a lot of uncertainty related to farming operations. Whether the field was treated evenly or with variable rate application, the performed field work is often inaccurate and vague. This is mainly because driving patterns in the field are not accurate and the optimal machinery input amount needs are uncertain. It is important to study this uncertainty in order to understand its impact on precision farming. This dissertation approaches the problem of uncertainty by applying geographical information quality evaluation and measurement methods to the evaluation of farm work execution. This makes it possible to exploit existing evaluation principles for spatial data in farm work execution cases. ISO 19157:2013 standard defines the key elements in spatial data quality. By studying agricultural technology research case studies, the effect of the quality elements and the estimated overall accuracy level of the farming work execution in grain production were constructed. The different case studies integrated remote sensing technologies such as satellite images and drones with cameras and hyperspectral technologies together with different spatial farm data and farmer's tacit knowledge. The case studies included also positioning errors based on dynamic GNSS positioning accuracy measurements and simulations and measured field work driving accuracies of farmers. The temporal quality was studied by developing methods how to apply real time data from external sources in ISOBUS environment. The machinery driving lines were overlapping by 10 % on average. Accurate steering assistance can cut that in half but there are still remarkable overlaps especially in headland areas. The biggest difficulty is the optimization of the variable rate application levels meaning the thematic accuracy. The thematic accuracy was determined as the variation of different tasks conducted for the same purpose being 22 % on average. The temporal accuracy was completely a case dependent containing a response to the immediate rain forecasts or applicability of one month old satellite image. A single precision farming operation was estimated to benefit about 31 €/ha which was estimated to be only 23 % of the total precision farming benefit potential according to the variables in this work. The overall accuracy of spatial data inputs was estimated to be 61 % in relation to optimal treatment in the studied cases. This number indicates the quality of spatial data inputs to farm machinery. This uncertainty is large in contrast to typical attempted precision farming adjustments and defined machinery performance requirements. These results suggest that there is a need for better uncertainty management, before different precision farming applications can truly be developed and evaluated.",

keywords = "spatial data quality, GNSS, UAV, agriculture, uncertainty, precision farming, paikkatiedon laatu, maatalous, t{\"a}sm{\"a}viljely, spatial data quality, GNSS, UAV, agriculture, uncertainty, precision farming",

author = "Jere Kaivosoja",

year = "2019",

language = "English",

isbn = "978-952-60-8840-2",

series = "Aalto University publication series DOCTORAL DISSERTATIONS",

publisher = "Aalto University",

number = "217",

address = "Finland",

school = "Aalto University",

}

TY - BOOK

T1 - Role of spatial data uncertainty in executions of precision farming operations

AU - Kaivosoja, Jere

PY - 2019

Y1 - 2019

N2 - The topic of this dissertation is the quality of spatial data related to arable farming work executions. There is a lot of uncertainty related to farming operations. Whether the field was treated evenly or with variable rate application, the performed field work is often inaccurate and vague. This is mainly because driving patterns in the field are not accurate and the optimal machinery input amount needs are uncertain. It is important to study this uncertainty in order to understand its impact on precision farming. This dissertation approaches the problem of uncertainty by applying geographical information quality evaluation and measurement methods to the evaluation of farm work execution. This makes it possible to exploit existing evaluation principles for spatial data in farm work execution cases. ISO 19157:2013 standard defines the key elements in spatial data quality. By studying agricultural technology research case studies, the effect of the quality elements and the estimated overall accuracy level of the farming work execution in grain production were constructed. The different case studies integrated remote sensing technologies such as satellite images and drones with cameras and hyperspectral technologies together with different spatial farm data and farmer's tacit knowledge. The case studies included also positioning errors based on dynamic GNSS positioning accuracy measurements and simulations and measured field work driving accuracies of farmers. The temporal quality was studied by developing methods how to apply real time data from external sources in ISOBUS environment. The machinery driving lines were overlapping by 10 % on average. Accurate steering assistance can cut that in half but there are still remarkable overlaps especially in headland areas. The biggest difficulty is the optimization of the variable rate application levels meaning the thematic accuracy. The thematic accuracy was determined as the variation of different tasks conducted for the same purpose being 22 % on average. The temporal accuracy was completely a case dependent containing a response to the immediate rain forecasts or applicability of one month old satellite image. A single precision farming operation was estimated to benefit about 31 €/ha which was estimated to be only 23 % of the total precision farming benefit potential according to the variables in this work. The overall accuracy of spatial data inputs was estimated to be 61 % in relation to optimal treatment in the studied cases. This number indicates the quality of spatial data inputs to farm machinery. This uncertainty is large in contrast to typical attempted precision farming adjustments and defined machinery performance requirements. These results suggest that there is a need for better uncertainty management, before different precision farming applications can truly be developed and evaluated.

AB - The topic of this dissertation is the quality of spatial data related to arable farming work executions. There is a lot of uncertainty related to farming operations. Whether the field was treated evenly or with variable rate application, the performed field work is often inaccurate and vague. This is mainly because driving patterns in the field are not accurate and the optimal machinery input amount needs are uncertain. It is important to study this uncertainty in order to understand its impact on precision farming. This dissertation approaches the problem of uncertainty by applying geographical information quality evaluation and measurement methods to the evaluation of farm work execution. This makes it possible to exploit existing evaluation principles for spatial data in farm work execution cases. ISO 19157:2013 standard defines the key elements in spatial data quality. By studying agricultural technology research case studies, the effect of the quality elements and the estimated overall accuracy level of the farming work execution in grain production were constructed. The different case studies integrated remote sensing technologies such as satellite images and drones with cameras and hyperspectral technologies together with different spatial farm data and farmer's tacit knowledge. The case studies included also positioning errors based on dynamic GNSS positioning accuracy measurements and simulations and measured field work driving accuracies of farmers. The temporal quality was studied by developing methods how to apply real time data from external sources in ISOBUS environment. The machinery driving lines were overlapping by 10 % on average. Accurate steering assistance can cut that in half but there are still remarkable overlaps especially in headland areas. The biggest difficulty is the optimization of the variable rate application levels meaning the thematic accuracy. The thematic accuracy was determined as the variation of different tasks conducted for the same purpose being 22 % on average. The temporal accuracy was completely a case dependent containing a response to the immediate rain forecasts or applicability of one month old satellite image. A single precision farming operation was estimated to benefit about 31 €/ha which was estimated to be only 23 % of the total precision farming benefit potential according to the variables in this work. The overall accuracy of spatial data inputs was estimated to be 61 % in relation to optimal treatment in the studied cases. This number indicates the quality of spatial data inputs to farm machinery. This uncertainty is large in contrast to typical attempted precision farming adjustments and defined machinery performance requirements. These results suggest that there is a need for better uncertainty management, before different precision farming applications can truly be developed and evaluated.

KW - spatial data quality

KW - GNSS

KW - UAV

KW - agriculture

KW - uncertainty

KW - precision farming

KW - paikkatiedon laatu

KW - maatalous

KW - täsmäviljely

KW - spatial data quality

KW - GNSS

KW - UAV

KW - agriculture

KW - uncertainty

KW - precision farming

M3 - Doctoral Thesis

SN - 978-952-60-8840-2

T3 - Aalto University publication series DOCTORAL DISSERTATIONS

PB - Aalto University

ER -

Role of spatial data uncertainty in executions of precision farming operations

Abstrakti

Tutkimusalat

Pääsy asiakirjaan

Sormenjälki

Siteeraa tätä