Agricultural Robot Project: Literature Review and Research Questions

This report provides a literature review on autonomous agricultural robots and analyzes research questions related to robust autonomy. It also demonstrates a theoretical method and an effective experimental setting-up along with illustrating the time schedule of the project through a Gantt Chart. Subject: Agriculture, Course Code: N/A, Course Name: N/A, College/University: N/A

Running head: AGRICULTURAL ROBOT PROJECT
Agricultural Robot Project
Name of the student:
Name of the university:
Author Note

1AGRICULTURAL ROBOT PROJECT
Executive summary
“Agbots” or agricultural robots are used in agriculture. Whenever any farm grows in size, they are
needed to automate various kinds of implementations. These are to be used over them. In this report,
a literature review is done and different research questions are analyzed keeping “autonomous
Agricultural Robot towards robust autonomy” in mind. It also demonstrates a theoretical method and
an effective experimental setting-up along with illustrating the time schedule of the project through a
Gantt Chart. As last, the outcomes and its various importances are also analyzed here.

2AGRICULTURAL ROBOT PROJECT
Table of Contents
1. Introduction:......................................................................................................................................3
2. Literature review on prior studies on autonomous agricultural robots:.............................................3
3. Research questions, aims and sub-goals:...........................................................................................8
3.1. Research questions:....................................................................................................................8
3.2. Research aims:............................................................................................................................8
3.3. Research sub-goals:....................................................................................................................9
4. Theoretical Content:..........................................................................................................................9
5. Experimental setup:.........................................................................................................................10
5.1. Discussion of field set-up and lab:............................................................................................10
5.2. Potential limitations:.................................................................................................................12
6. Results, outcomes and relevance:....................................................................................................13
7. Project planning and Gantt chart:....................................................................................................13
7. Conclusion:......................................................................................................................................17
8. References:......................................................................................................................................18

3AGRICULTURAL ROBOT PROJECT
1. Introduction:
The agricultural robots or “Agbots” are implemented for agriculture. As any farm grows in
size, they with the volume of applications used over them have been needed for ways of automating
them.
This was done manually before. Now, the tasks are performed by those autonomous
machines, as they need multiple repetitions over a large area and an extended period. The usage of
agricultural robots is designed as standard types of equipment for farms. This includes pesticide
sprayers, various combines and tractors.
The following study has conducted a literature review and analyzed various research
questions from “autonomous Agricultural Robot towards robust autonomy” from the mechanical
domain. Then the theoretical methodology and experimental set-up are demonstrated. Lastly, the
results and its relevance are discussed.
2. Literature review on prior studies on autonomous agricultural robots:
It is seen that in most of the cases, robots are ineffective at doing farming jobs. This has
commonly needed vast amounts of materials like fertilizers and seeds or is retrieved from the
harvesting field. This has been dealing with mapping and controlling precision and field for spraying
pesticides. It takes place because of low weight in comparison to a tractor that makes minor soil
compaction. Here, the degree of soil compaction is vital to consider mapping and monitoring that
which is often done numerous times in a year. Oberti et al. (2016) explains that this is because soil
compaction has been causing various issues like a decrease in denitrification and crop growth.
Agrobots have been altering the scenario of agriculture beyond identification. This is from robot-

4AGRICULTURAL ROBOT PROJECT
assisted milking to several types of cow-herding drones. The food industry gets revolutionised
through automation and robotics. The mechanised agriculture has to move as per strength. These are
actual issues in modern agriculture. Conventional methods of farming are to keep up the impacts
needed by the current market. The farmers in the first-world countries have been suffering from the
lack of workforce. Here, the rise of automated farming has been an attempt to resolve the issues
through advanced and robotic sensing. As per the current report, shown by Lopes et al. (2016) the
marketplace for agricultural drones and robots are intended to reach about 30 billion dollars for the
upcoming five years. Further, there are various issues in modern agriculture.
Conventional methods have been struggling with keeping up the activities needed by the
business. For example Ball et al. (2016) discussed that there is a rise in demand for nursery
automation. Organizations such as “HETO Harvest Automation and Agrotechnics” has been
delivering solutions regarding warehousing, potting and seeding living plants are greenhouses. The
autonomous precession seeding has been assimilated robotics with geomapping. The map is
generated by Bogue (2016) in his article showing soil properties like density and quality at all kinds
of points at the field. Different drone companies have been offering farmers with various combined
packages including robotic hardware and software of analysis. Farmers have been moving drones to
the pastures initiating software through smartphone or tablet and see that collected crop information
in real time. Again various ground-based robots have been supplying more detailed controlling as
they get closer to crops. Few of them have been using activities like fertilising and weeding.

5AGRICULTURAL ROBOT PROJECT
ATMEGA
16
434 MHz
Receiver
Decoder
Solar
Panel
Power
Supply
Battery
Motor
Driver
(L293D)
DC Motor
Motor
Driver
(PSNF06)
DC Motor
Figure 1: “Basic Block Diagram of Agriculture Robot”
(Source: Dunlop, 2018)
According to Wable, Khapre and Mulajkar (2016), robots have the benefits as they access
sectors where the machines have not been able to do. Here, for instance, growers of corns have been
facing challenges that the plants have been growing fast to fertilise them dependably. The robots
have aimed to resolve issues efficiently driving the rows of many corns. This has also been included
nitrogen fertiliser directly as the ground of every plant. The ides of micro-spraying as discussed by
Bloch, Bechar and Degani (2017) has decreased the quantity of herbicide utilised in growing crops.
The micro-spraying robots have used the technology of computer vision for detecting weeds and
spraying targeted drops of herbicides. AG BOT II has been a solar-powered machine using that kind
of processes. Mueller-Sim et al. (2017) mentioned about a LettuceBot thinning robot that has
achieved the award for outstanding innovation for agriculture. It helps in deciding at what time the

6AGRICULTURAL ROBOT PROJECT
plants are to be kept and what to eradicate. There has been a rise in trend for followers that lead the
autonomy. Here, the tractors have been following a human-driven combination of harvesters
autonomously for collecting grains. For instance, EU-funded smart robots for crop projects have
been progressing on various harvesting applications involving apple harvesting, sweet pepper-
picking and grape picking.
Current projects:
The EU-funded "Clever Robots for Crops" project is making progress on few harvesting
applications, including apple harvesting, grape picking and sweet pepper picking as shown by
Jasiński et al. (2018). Though maximum of the agricultural robots has been applied in growing of
crops, there has been emerging applications under cattle and sheep farming. This is done through
assimilating capabilities of an aerial survey of various small autonomous UAV or Unmanned Aerial
Vehicle multi-copter with different agricultural unmanned ground vehicles. Here the system has
been surveying the field from the air and performing a targeted intervention at the ground. They
have been delivering in-depth data regarding decision support with minimal invasion of users. Here
the system has been adapting a broad range of crops though choosing various ground level treatment
packages and sensors. Here the development has needed developments in abilities of technology for
secure and exact navigations under farms. This is coordinated through multi-robot mission planning
enabling a full field of survey. It has also included multispectral mapping, as demonstrated by
Zakaria (2017), which is three-dimensional mapping having spatial resolution and high temporal
interventions of techniques and tools. It has also included tools of data analysis from weed detection
and crop monitoring and design of user interface for supporting division making at agriculture.
Besides, a gap has been present as mentioned in the artcle of Al-Beeshi et al. (2015),
stopping the effective transition from scientific to societal and economic effect. This is also referred

7AGRICULTURAL ROBOT PROJECT
to as “Technological Innovation Gap”. EU-FP7-project-wide research is done on agricultural
robotics. Here an application has been “Sweet pepper harvesting robot”. This kind of robot has been
technically and economically viable as demonstrated by Baxter et al. (2018). The software and
hardware models proven have developed crops used as groundwork. These successful “CROPS”
software module has been from ROS or “Robotic-Operating-System” expanded and maintained at
SWEEPER. Further, the gripper and effectors have been retained. Here, the patent-pending module
has been grasping sweet pepper instead of any necessity of proper measurement of orientation and
position of fruits. At SWEEPER, the system of cropping is intended to optimise for facilitating
harvesting of robots. It was concluded at CROPS that instead of any 9DOF, one 4DOF arm has been
enough to decrease the expenses. As per Ulbrich et al. (2015), for developing abilities level of
abilities regarding cognitive capabilities, the plant models would be deployed for proper location of
sweet peppers. Here, the model-based vision has been increasing and quickening up detection of
fruits. Based on various insights of CROPS, the sensors have been placed only over the gripper.
Further, LightField sensor has been introduced that has been able o record colour and 3D
information at the same time.
Factoids to be considered for the research:
It is seen that most of the conglomerates of US farming have been buying various foreign
areas and starting a farm there through citing the overall lesser expenses. For instance, Robert (2017)
investigated that China has been buying land at Africa and sending expert workers for supervising
those farms. International ranchers and farmers have been transitioning to precision methods of
agriculture. This has indicated the dividing the acreage into various sub-plots, in multiple cases,
proper down to distinct flora and fauna has been helping to the rise the productivity and decreasing
entire costs. Different unmanned aerial vehicles can be utilised for spraying, sensing observing and

8AGRICULTURAL ROBOT PROJECT
mapping. Autonomous or unmanned ground vehicles have been supplying more precise movements
and thus helping with precision practices. It is seen from the article of Baxter et al. (2018), the report
of US Bureau of Labor Statistics that 2012 median pay for the farm workers have been about 9
dollars. On the other hand the reports from US Bureau of Labor Statistics that has been about
750,000 agricultural workers in 2012 that has been down about 3% from 2011. The approximate
number of crop workers has been 74% in US-born in Central America and Mexico where more than
half has been still kept undocumented as per Fortune Magazine. Furthermore, Durmuş et al. (2015)
analyzed that Cropdusters has been possessing 3rd largest fatality rate taking place among
professionals at the U.S. Here, 90% of crop saying at Japan has been done through different
unmanned helicopters. Again, ResearchMoz has projected that the size of agricultural marker would
get increased from about 817 million dollars to about 16 billion dollars from 2013, till the end of
2020 as shown by Wable, Khapre and Mulajkar (2016).
3. Research questions, aims and sub-goals:
3.1. Research questions:
 What is the status of present trends and deployments of free and agricultural patterns?
 What is the potential of future applications for autonomous agricultural robots?
 How are these autonomous vehicles different from those of conventional ones?
 What are the field operations for crop establishments, plant cares and selective harvestings?
3.2. Research aims:
The various purposes include the following. Example of this includes driving in top rows for
a maximum of 30 seconds. Further, the robot has been continuing with operation with the fault of

9AGRICULTURAL ROBOT PROJECT
sensor and actuator. Also, operating under faulty and normal activities has been highly dangerous for
the environment.
3.3. Research sub-goals:
The sub-goals of robots in agriculture have been immense. The robots were appearing at
farms in different guises and rise in numbers. The various issues related to free farm tools have been
overcoming the tools. The device would turn out to be the future, and there have been essential
causes to think that it has not been replacing the human driver with computers. This has indicated
rethinking of how the product can be done. The production of crops has been cheaper and better with
a swarm of few machines than various large ones.
4. Theoretical Content:
The agriculture industry has been under transition. The transition has been differing as per
the country, states, and regions and practiced by farming. This has taken place from primitive o
traditional and from precession to experimental. This little bit of everything has been going on at
every position. However, any general trend worldwide has been towards the precision agriculture
that has been supplemented though developed technologies that have included robotics (Wang et al.
2016).
Various factors have been precipitating within those changes apart from an international
growth of populations and availability and cost of the labor. This has included a decrease in
availability and rise in an expense of water, political and processes that are regulatory. It has also
included restricted tillable acreages, cheaper, better and quicker technological automation resources
and changes in climate (Duarte et al. 2016). Current ranchers and farmers have been highly
technical. Different digitally controlled farm deployments have been in use regularly.

10AGRICULTURAL ROBOT PROJECT
There have been numerous automatic devices present for various elements of functions
related to agriculture. This has extended from grafting to planting. This has included packaging to
boxing, harvesting to sorting. The farmers can use software systems and various maps of aerial
surveys and information for guiding the field operations (Serrano et al. 2017). Moreover, they have
been using auto-steer systems including different new tractors following GPS and guidance of
software. Besides, many farmers have been transitioning few operations to total autonomy. In this
way, forward-thinking owners for farms of the current age has been able o skip the over slow with
various developed improvements and then directly jump to autonomous and robotic automation
(Jasiński et al. 2018).
5. Experimental setup:
5.1. Discussion of field set-up and lab:
Fault analysis is to be performed with severity analysis of every wheel, proximity sensor and
inclinometer faults. Next, a non-linear model is to be implemented and designed from FDI method
or Fault Detection and Isolation method (Grimstad and From 2017). Next, the critical errors are to be
verified. For complementing that linear method two more new ways for FDI have been examined,
that has been resulting in various objectives. The first one is to implement and design non-linear FDI
method. Further, the practical goals have included extra software and hardware (Radkowski 2018).
Next, an implementation and designing of proximity sensors are to be implemented on that API.
After this, a space inclinometer has been performed and designed for providing rolling and pitching
measurements. Lastly, implementing and developing of relays are included for disconnecting distinct
wheels.

11AGRICULTURAL ROBOT PROJECT
Sensors and inputs Sate 1: Kalman Filter
State 2: Kalman Filter
State n Kalman Filter
Sate Observer
Discrete
Sate
Estimate q
Figure 2: “Structure of Modified Multiple Hypothesis Test State Observer”
(Source: Hsu and Hsu, 2018)

12AGRICULTURAL ROBOT PROJECT
Figure 3: “The placement of the Proximity Supervisor in the control structure of the API”
(Source: Popular Mechanics, 2018)
5.2. Potential limitations:
The various restrictions are listed hereafter.
 There must not be any interference with or disability of emergency stop buttons that is
mounted over to the autonomous robots.
 They must be no providing to wheels even though the OBC gets shut down.
 They must be able to turn the power of wheels as it gets on or off.
 The wheels might turn off, through using various voltages supplied by the parallel port over
that OBC.

13AGRICULTURAL ROBOT PROJECT
6. Results, outcomes and relevance:
As the robot operates nominally and driving to that field, the planting with sensitive crops or
with crops has been containing enough width taking place between the rows. This has been fitting
the wheels of that robot driving on the topmost area of the plants. However, as any error take place,
the robot can get deviated from the ordinary course. This, for example, includes driving of top rows
for the utmost of 30 seconds. Further, the autonomous robots have been conducting operations under
a sensor or actuator fault. Here, the robots have been kept operational under erroneous situations as
long as it has been possible. Various processes within faulty and normal operations have not been
effectively harmful to the scenario.
7. Project planning and Gantt chart:
Task Name Duration Start Finish
Fault Detection and isolation 97 days Mon
5/21/18
Tue
10/2/18
Fault Analysis 92 days Mon
5/21/18
Fri 9/28/18
Isolability Analysis 2 days Mon
10/1/18
Tue
10/2/18
Accept Test of Linear FDI 192 days Wed
10/3/18
Fri 7/5/19
Accept Test of particle filter-FDI method 110 days Mon
7/8/19
Fri 12/6/19
Accept Test of Active Fault Isolation Supervisor 2 days Mon Tue

14AGRICULTURAL ROBOT PROJECT
12/9/19 12/10/19
Test of steering fault isolation 1 day Mon
12/9/19
Mon
12/9/19
Test of propulsion Fault isolation 1 day Tue
12/10/19
Tue
12/10/19
Hardware Test 5 days Wed
12/11/19
Tue
12/17/19
Inclinometer Test 3 days Wed
12/11/19
Fri
12/13/19
Proximity sensor test 2 days Mon
12/16/19
Tue
12/17/19
Implementing Software 2 days Wed
12/18/19
Thu
12/19/19
Simulink blocks 1 day Wed
12/18/19
Wed
12/18/19
Stabdalone Programs 1 day Thu
12/19/19
Thu
12/19/19
FDI method test 2 days Fri
12/20/19
Mon
12/23/19
Test of UIO method 1 day Fri
12/20/19
Fri
12/20/19
Test of Beard Fault detection Filter method 1 day Mon
12/23/19
Mon
12/23/19
Active FI Supervisor 7 days Tue Wed

15AGRICULTURAL ROBOT PROJECT
12/24/19 1/1/20
Active Isolation of steering faults 2 days Tue
12/24/19
Wed
12/25/19
Active isolation of propulsion faults 1 day Thu
12/26/19
Thu
12/26/19
Kickoff Meeting: Starting Projects Right 4 days Fri
12/27/19
Wed
1/1/20
Getting the client on-side 1 day Fri
12/27/19
Fri
12/27/19
Approval process – the process and personnel
for signing off deliverables
1 day Mon
12/30/19
Mon
12/30/19
SoW Review 1 day Tue
12/31/19
Tue
12/31/19
Identifying RAID (Risks, Assumptions, Issues,
Dependencies) and change management
1 day Wed
1/1/20
Wed
1/1/20

16AGRICULTURAL ROBOT PROJECT
Global Music
Festival_final
Fault Detection and
isolation
Fault Analysis
Isolability Analysis
Acceptest of Linear
FDI
Accept Test of
partcile filter-FDI
method
Accept Test of Active Fault
Isolation Supervisor
Test of steering fault
isolation
Test of propulsion
Fault isolation
Hardware Test
Inclinometer Test
Proximity sensor
test
Implementing Software
Simulink blocks
Stabdalone
Programs
FDI method test
Test of UIO method
Test of Beard Fault
detection Filter
method
Active FI Supervisor
Active Isolation of
steering faults
Active isolation of
propulsion faults
Kickoff Meeting: Starting
Projects Right
Getting the client
on-side
Approval process –
the process and
personnel for signing
off deliverables
SoW Review
Identifying RAID
(Risks, Assumptions,
Issues,
Dependencies) and
change management
Figure 4: “Time schedule for current Agricultural Robot Project”
(Source: Created by author)

17AGRICULTURAL ROBOT PROJECT
7. Conclusion:
The above discussion has highlighted the vision of how different elements of crop production
have been turning to an automated one. Though the current manned operations have been active over
huge sets, there have been sectors to reduce the treatment scaled having autonomous machines. This
has resulted in high efficiencies. This process of developments has been incremental. However, the
entire idea has needed a “paradigm shift”. This has taken place in the way in which mechanization
for crop production has been based on the needs of plants. Moreover, this has needed a novel
approach to meet them instead of altering the current applications. Again, at modern greenhouses,
there has been a rise in demand for automated labours. Here, the availability of skilled workforces
accepting repetitive activities under adverse climatic conditions of the greenhouse has been declining
fast. Here the rise in labour costs has decreased of capacity has put extraordinary pressure over the
competitiveness of the sector of European conservatory. The study shows that the current
robotisation of the labour has entered to a considerable level of readiness concerning technology.

18AGRICULTURAL ROBOT PROJECT
8. References:
Al-Beeshi, B., Al-Mesbah, B., Al-Dosari, S. and El-Abd, M., 2015, May. iplant: The greenhouse
robot. In Electrical and Computer Engineering (CCECE), 2015 IEEE 28th Canadian Conference
on (pp. 1489-1494). IEEE.
Ball, D., Upcroft, B., Wyeth, G., Corke, P., English, A., Ross, P., Patten, T., Fitch, R., Sukkarieh, S.
and Bate, A., 2016. Vision‐based Obstacle Detection and Navigation for an Agricultural
Robot. Journal of Field Robotics, 33(8), pp.1107-1130.
Baxter, P., Cielniak, G., Hanheide, M. and From, P., 2018, March. Safe Human-Robot Interaction in
Agriculture. In Companion of the 2018 ACM/IEEE International Conference on Human-Robot
Interaction (pp. 59-60). ACM.
Bergerman, M., Maeta, S.M., Zhang, J., Freitas, G.M., Hamner, B., Singh, S. and Kantor, G., 2015.
Robot farmers: Autonomous orchard vehicles help tree fruit production. IEEE Robotics &
Automation Magazine, 22(1), pp.54-63.
Bloch, V., Bechar, A. and Degani, A., 2017. Development of an environment characterization
methodology for optimal design of an agricultural robot. Industrial Robot: An International
Journal, 44(1), pp.94-103.
Bogue, R., 2016. Robots poised to revolutionise agriculture. Industrial Robot: An International
Journal, 43(5), pp.450-456.
Duarte, M., dos Santos, F.N., Sousa, A. and Morais, R., 2016. Agricultural wireless sensor mapping
for robot localization. In Robot 2015: Second Iberian Robotics Conference (pp. 359-370). Springer,
Cham.

19AGRICULTURAL ROBOT PROJECT
Dunlop, T. (2018). Agbots, next gen farming and how they can teach us about the future of work.
[online] the Guardian. Available at:
https://www.theguardian.com/sustainable-business/2017/may/09/agbots-next-gen-farming-and-how-
they-can-teach-us-about-the-future-of-work [Accessed 23 Jun. 2018].
Durmuş, H., Güneş, E.O., Kırcı, M. and Üstündağ, B.B., 2015, July. The design of general purpose
autonomous agricultural mobile-robot:“AGROBOT”. In Agro-Geoinformatics (Agro-
geoinformatics), 2015 Fourth International Conference on (pp. 49-53). IEEE.
Grimstad, L. and From, P.J., 2017. Thorvald ii-a modular and re-configurable agricultural
robot. IFAC-PapersOnLine, 50(1), pp.4588-4593.
Hsu, J. and Hsu, J. (2018). Rise of the Ag-Bots Will Not Sow Seeds of Unemployment. [online]
Scientific American. Available at: https://www.scientificamerican.com/article/rise-of-the-ag-bots-
will-not-sow-seeds-of-unemployment/ [Accessed 23 Jun. 2018].
Jasiński, M., Mączak, J., Szulim, P. and Radkowski, S., 2018, March. Autonomous Agricultural
Robot–Testing of the Vision System for Plants/Weed Classification. In Conference on
Automation (pp. 473-482). Springer, Cham.
Lopes, C.M., Graça, J., Sastre, J., Reyes, M., Guzmán, R., Braga, R., Monteiro, A. and Pinto, P.A.,
2016. Vineyard yeld estimation by VINBOT robot-preliminary results with the white variety
Viosinho. In Proceedings 11th Int. Terroir Congress. Jones, G. and Doran, N.(eds.), pp. 458-463.
Southern Oregon University, Ashland, USA.. Jones, G.; Doran, N.(eds.).
Mueller-Sim, T., Jenkins, M., Abel, J. and Kantor, G., 2017, May. The Robotanist: a ground-based
agricultural robot for high-throughput crop phenotyping. In Robotics and Automation (ICRA), 2017
IEEE International Conference on(pp. 3634-3639). IEEE.

20AGRICULTURAL ROBOT PROJECT
Oberti, R., Marchi, M., Tirelli, P., Calcante, A., Iriti, M., Tona, E., Hočevar, M., Baur, J., Pfaff, J.,
Schütz, C. and Ulbrich, H., 2016. Selective spraying of grapevines for disease control using a
modular agricultural robot. Biosystems Engineering, 146, pp.203-215.
Popular Mechanics. (2018). 5 Agro-Bots That Will Change How We Grow Everything. [online]
Available at: https://www.popularmechanics.com/technology/robots/g1867/5-farm-robots-ag-
industry/ [Accessed 23 Jun. 2018].
Radkowski, S., 2018. Autonomous Agricultural Robot–Testing of the Vision System for
Plants/Weed Classification. Automation 2018: Advances in Automation, Robotics and Measurement
Techniques, 743, p.473.
Robert, C., 2017, October. First Insights into Testing Autonomous Robot in Virtual Worlds. In 2017
IEEE International Symposium on Software Reliability Engineering Workshops (ISSREW) (pp. 112-
115). IEEE.
Serrano, D., Astolfi, P., Bardaro, G., Gabrielli, A., Bascetta, L. and Matteucci, M., 2017, November.
GRAPE: Ground Robot for vineyArd Monitoring and ProtEction. In ROBOT 2017: Third Iberian
Robotics Conference (Vol. 1, p. 249). Springer.
Ulbrich, H., Baur, J., Pfaff, J. and Schuetz, C., 2015. Design and realization of a redundant modular
multipurpose agricultural robot. In Proceedings of the XVII International Symposium on Dynamic
Problems of Mechanics (DINAME), Natal, Brazil.
Wable, A.A., Khapre, G.P. and Mulajkar, R.M., 2016. Intelligent Farming Robot for Plant Health
Detection using Image Processing and Sensing Device. International Journal of Engineering
Science, 8320.

21AGRICULTURAL ROBOT PROJECT
Wang, Z., Gong, L., Chen, Q., Li, Y., Liu, C. and Huang, Y., 2016, August. Rapid Developing the
Simulation and Control Systems for a Multifunctional Autonomous Agricultural Robot with ROS.
In International Conference on Intelligent Robotics and Applications (pp. 26-39). Springer, Cham.
Zakaria, R.N.B., 2017. Design of UV-Bio configuration of the NMBU agricultural robot (Master's
thesis, Norwegian University of Life Sciences, Ås).