Ledige phd-stillinger til innovasjonsprosjekt

ITS søker fire kandidater innen fagfeltene autonome systemer og solcelleteknologi. 


Autonomous monitoring, control and protection of renewable energy infrastructure. “Innovation project” with four PhD positions

Department of Technology Systems (ITS), University of Oslo (UiO)




The integration of numerous renewable energy sources into the power grid will require careful monitoring and control of the energy production infrastructure. This project deals with external and internal surveillance and monitoring of solar power plants, for autonomous control, protection and steering of the energy production. Autonomous drones with onboard cameras will provide information of external power plant parameters, which shall be correlated with internal solar cell parameters, other available data and computer models of the power plants. A demonstrator shall be built and tested on an existing industrial facility, as an essential part of a fourth “innovation year” of each PhD project. The project has considerable potential for innovation, in that it addresses issues that are relevant already today, and involves commercial operators.

Utlysningen finner du her




The project is motivated by two anticipated transformations of society: the transition to distributed, renewable and smart energy systems, as well as the introduction of autonomous systems in large numbers.


Tomorrow's energy infrastructure will include large numbers of distributed systems and components, such as small and large renewable energy production facilities, energy storage facilities, transport infrastructure etc. In this project we will focus on monitoring of photovoltaic power systems in particular, but the methodology may also be transferred to other types of infrastructure. The solar cell industry is in a very exciting phase, with solar cell panels rolling out in huge volumes with nearly 500 km2 per year at current pace. In 2016, an estimated 70 GW(peak) of such solar cell facilities are installed at an investment cost of about 1000 billion NOK. The systems installed are either large solar parks, often with hundreds of thousands of solar panels, or smaller systems on existing or new buildings, sometimes with a few tens of panels, and which may be fairly inaccessible.


The monitoring and operation of such energy infrastructure will be very demanding and, potentially, very costly. Especially for the smaller power plants the cost of instrumentation and monitoring could be disproportionally large if instrumentation and/or hourly costs for personnel are included. Automatic remote monitoring, control and protection of both large and small systems are therefore highly desirable. Autonomous systems open up new cost-efficient operations of such systems without the presence of physical operators.


This project also aims to be a platform for the development of related projects in renewable energy systems and within autonomous systems. In the short term, the goal is to develop new projects with industry supported by the Norwegian Research Council (NRC). In the long term, the goal is to participate in Horizon2020, NRC funded Centers of Excellence, as well as more basic research.


About the project


Today there is an increasing demand for drones for monitoring of photovoltaic power plants. In this project we will promote a long term development plan, which we suppose will take place over three generations. In today's generation, remote controlled drones combined with electronic readout of key production and measurement parameters are the basis for monitoring and decisions making. This is a process highly dependent on manual operation. In the next generation, we envisage the use of single, autonomous drones that enable more effective monitoring, mainly for larger photovoltaic power plants. In the third generation, we envision a greater number of interacting autonomous drones. This stage is expected to be important for the monitoring of the large number of small facilities in cities, where each system alone will be too small to justify the purchase of surveillance equipment.


The Institute for Energy Technology (IFE) has access to several larger and smaller facilities across the country, and a large power plant in South Africa. The availability of these power plants assure access to all kinds of data that are relevant for this project. However, we also believe that the proposed methodology can be used much broader within the energy industry. Autonomous systems should be able to monitor internal and external performance parameters in order to ensure robust production of electricity. An important part of this is mobile autonomous drones that can fly over the solar panels and detect faults and anomalies. This network of drones must collaborate autonomously to decide where the drones’ presence is needed and which task the drones should prioritize. Based on data from these drones and internal performance parameters, the system should be able to decide and implement actions in order to maintain production, both in terms of reprogramming the system parameters and/or dispatching other unmanned vehicles that physically correct the error.


In addition to strengthening the activities on energy systems and autonomous systems at ITS, the project could also be a platform for a much broader cooperation. Navigation and communication will be key components within research in autonomous system at ITS, along with autonomy research at the Norwegian Defence Research Establishment (FFI). Furthermore, the project opens opportunities for interaction and cooperation in other areas like advanced signal processing, machine learning, artificial intelligence, measurement systems, sensor technology and relevant topics in energy science. Research in these disciplines is also addressed by research groups at the Department of Computer Science and Physics at UiO, at UiO Blindern.




This project will be implemented in an exciting cooperation between IFE, FFI and UiO. The theme is highly relevant for large ongoing international research activities, both within the monitoring and operation of renewable power plants and autonomous systems. The rapid pace of development in these areas will most likely benefit the Norwegian society and industry in years to come. There are a number of exciting industrial partners in Norway within these research areas. FFI and IFE have good contacts with important partners that already have significant commercial interest related to these activities. Consequently, there are good opportunities for innovation projects within existing businesses areas, as well as the possibility of generating novel concepts and solutions aimed at the development of new companies with potential for long-term growth.




The forth "innovation year" for each PhD will be devoted to the realization of a demonstrator. This is to highlighting the innovation potential of the research and to make the 4 PhDs work closely together. A demonstrator will also provide good visibility. It will also raise the awareness of the research related ambitions and promote cooperation within related research areas at ITS, FFI, FFI and with important, relevant departments at UiO, particularly the Departments of Informatics, Physics, Chemistry and Mathematics. A demonstrator will also facilitate technology transfer from research to industrialization.


We propose an ambitious demonstrator that will take the research well beyond what can be termed as "state-of-the-art" today. We propose a demonstration where a swarm of drones (3+) monitors an instrumented solar park and based on data abstracted from internal and external operational parameters continuously optimizes the energy output. This means fully autonomous monitoring and operation of a solar energy system without human intervention. We want to develop and test equipment at Kjeller in preparation for a full scale demonstration using a real industrial plant. A possible relevant plant could be Asko at Vestby with 10.000 to 20.000 panels.




The four PhDs will be co-located at ITS, and the project will be led by a project manager at ITS. The four PhD positions will cooperate with each other and with leading researchers at IFE and FFI, as well as relevant groups at UiO Blindern. Dedicated project meetings with all four PhD positions and their research groups will be held in circulation between ITS, IFE, FFI and UiO at Blindern.


There is also a goal of including a number of M.Sc. students in the project. This will create a broader and more dynamic community with new candidates to fill future PhD positions. This will increase the likelihood of success, especially in the long term beyond this project period.


Infrastructure and support for the continuous operation of a swarm of drones will be done in collaboration with FFI, and the availability of an instrumented solar park will be closely worked out with IFE.


Description of the tasks


Reliable navigation is a prerequisite for autonomous mobile robots. In this scenario the drones need robust navigation to confidently locate and align the sensors over the solar panels (PhD 1). Communication between the drones is needed in order for them to organize themselves and collectively perform the given task, and communication back to a ground station may provide real-time sensor data to the power plant controller. (PhD 2). Inner performance parameters of solar panels will be compared with sensor data from drones and other available data (PhD 3). Aggregated data will be relayed to a control center where they will continuously be analyzed before decisions on relevant measures will be taken by the system itself (PhD 4).


The autonomy in this scenario comprises several aspects: autonomous mobile robots (drones), communication networks, and autonomous decision making systems (center) that continuously work together to achieve the system collective goal of robust and cost-effective power production over time.



PhD 1: Robust navigation using ad hoc radio beacons


PhD 1 will study concepts for robust navigation that can also handle the loss of satellite signals. The main idea here is to generate and/or utilize existing RF sources as navigation beacons using software defined radio (SDR) as support for traditional navigation.





Navigation is required in all operations of mobile autonomous systems. In the case of such systems used for protecting critical energy infrastructure, navigation needs to be robust and reliable in all situations. Outdoor global navigation is usually based on satellite navigation like Global Positioning System (GPS). Missing or severely degraded satellite signals can be a problem for many navigation systems, and can put unmanned surveillance systems inoperable. These problems may be due to environmental factors as terrain shielding of satellite signals, multipath interference, atmospheric disturbances and accidental/intentional disruption of signals. Global Navigation Satellite Systems (GNSS) are not initially designed to work in the presence of interference sources. Satellite signals, which are attenuated below noise level of receiver when reaching the Earth's surface, are therefore very sensitive to inference. At the same time, studies in several countries, including Norway, conclude that interference, both accidental and intentional (jamming), is a growing problem. The increase in GNSS based monitoring in transport and service industry, tolling based on GNSS estimated travel distances, anti-theft and anti-terrorism surveillance, are all examples of activities that are vulnerable to the proliferation of inexpensive and readily available GNSS jammers. There are also indications that the access and availability of GNSS jamming/spoofing equipment will increase in the future. This will pose a threat to critical infrastructure that uses GNSS positioning and timing, creating a need for alternatives to satellite based navigation and time synchronization.


Description of task


The aim of this study is to investigate and examine the application of modern RF systems to increase the robustness and precision in outdoor navigation when access to satellite navigation is limited or absent. This work should be carried out in cooperation with researchers in cybernetics at ITS and researchers in autonomy at FFI. Ad hoc radio beacons that are placed in strategical positions can provide accurate navigation on local scale when satellite based navigation is impossible or insufficient. In addition, support from other navigation sensors such as inertial sensors, magnetometer, optical sensors (camera), and the ability to utilize other known and available RF sources (so called "Signals of Opportunity"), should be examined. Other GNSS, like the new European Galileo, should also be considered, especially with regards to encryption of signals in order to prevent spoofing. Modern SDRs are cheap, easy and flexible RF sources that may be suitable as ad hoc radio beacons, including onboard autonomous systems. SDRs can also be used as pure GNSS receivers or in a more complex navigation system that processes and merges data from different navigation sensors. This flexibility makes SDR an ideal platform for testing various navigation concepts.


An important part of the investigation is to study the possibilities and limitations of restoring navigation using a network of SDRs. Suitable waveforms for robust navigation must be studied. The PhD candidate should also consider various concepts for autonomous deployment of netted beacons. Another interesting area of research is to investigate multi-function systems, where the same SDR may have the role as both a beacon and navigation sensor. This would be an area where cooperation with PhD 2 is necessary.



PhD 2: Wireless communication in a system of mobile autonomous units


PhD 2 will study new concepts for wireless communication between mobile autonomous vehicles without fixed infrastructure.



Description of tasks


The task is to study wireless communication in a system of autonomous mobile units. Focus is to develop and test different concepts for communication between Unmanned Aerial Vehicles (UAV) as mobile autonomous nodes. Unmanned Ground Vehicles (UGV) and Unmanned Surface Vehicles (USV) may also be included as mobile autonomous nodes when relevant.


The aim of the study will be to achieve optimal connectivity in terms of the scarce communication resources available in autonomous mobile systems. The main application of this research is communication of UAV based sensor data from various types of photovoltaic power plants to some central data processing node on the ground.


In order for an autonomous system consisting of many units to solve a task collectively, the system must be able to communicate internally and often externally to other nodes as part of the mission. Self-configuring mobile communications networks have been researched for many years within the ad hoc network community. Focus has been on developing network protocols that provide the best possible communication services given node location and movement patterns. However, with the ability to place UAVs in 3 dimensions, novel network protocols should also include the dynamical positioning of the nodes for optimal communication. The choice of node location and radio communication carriers should be part of the system's autonomy, meeting the requirement for robustness, capacity and availability when adapting to the environment and the system goals.


Ad hoc communication could also be important in connection with robust navigation in a system of mobile autonomous units (ref PhD 1).


PhD 3: Measurement and imaging technology for condition monitoring and diagnostics of solar panels.





The main objective of this position is to develop measurement methods for precise condition monitoring and diagnostics for solar power plants in the field. Measurements of the actual performance of solar panels over time will be compared with theoretical prediction based on local meteorological measurements. Methods for robust determination of degradation mechanisms in solar modules will be developed based on imaging technology.


A solar plant's capacity at the time of installation is well defined. However, some decrease in performance over time is expected, partly because of material related degradation in the solar panels over time. This should be small (this is quantified in the various warranty documents), but errors can occur, either suddenly or over time, which significantly reduces the electricity production of a panel. Different degradation mechanisms will affect the electrical response of the solar panel. However, several types of degradation of different importance can provide similar changes in the electrical output of the solar panels. Fortunately, many of the faults can detected or identified by visual inspection or imaging, for example by well-known signatures in IR, either by electroluminescence (EL) or thermography. To detect and identify defected components will allow one to identify errors before they have large consequences for photovoltaic power plants' production or reliability. By developing tools for diagnostics that enable robust and cost-efficient recognition of degradation, as well as their priority, one can ensure reasonable maintenance of solar power plants and in principle extend their technical life far beyond the warranty period. This could have major beneficial economic and environmental consequences. Owners of solar installations will want to ensure good, but cost-effective maintenance, but also to identify any breach of the warranty of the system vendor. A system vendor will therefore also benefit from being able to identify defects in order to take the matter further with component suppliers.


Description of task


In this thesis, the status determination and diagnostics of solar panels is the main topic. The goal is to determine the so-called performance ratio, efficiency and degradation over time using a combination of measurements on solar panels and inverters, together with selected meteorological data such as temperature and insolation. Data from both smaller and larger power plants and power plants with various solar cell technologies will be analyzed. Performance under Norwegian and Nordic conditions will receive special interest since the public data on such systems remains limited. The measurements described above will be compared with information from imaging panels, initially with handheld tools, but over time using flying drones equipped with optical and / or infrared sensors. Initially, a commercially available remote-controlled drone will be used, but the goal is to eventually lay the groundwork for the development of autonomous surveillance systems for photovoltaic power plants in close cooperation with PhD 1 and PhD 2.


The candidate will have access to selected solar parks of larger and smaller size, primarily in Norway. Study and validation of methods for condition monitoring will be implemented on well instrumented power plants. However, most power plants are not instrumented for detailed condition monitoring. Therefore, it is an important goal to develop a methodology for connecting spread measurement data from a rapidly growing number of solar parks with different instrumentation as a source for detecting and identifying discrepancies. Such methodology may in a future with a great penetration of photovoltaic power plants be a robust method for identifying photovoltaic power plants with abnormal performance and degradation. The measured performance degradation will be correlated with information from optical and / or infrared cameras and sensors. This type of imaging has proven suitable to detect critical defects. Finally, the direct identification of the specific material related degradation mechanisms in a solar panel will be devoted interest. Experience with a remote-controlled drone will be a critical foundation to get a realistic picture of the possibilities of autonomous drones in the next generations of implementation.



PhD 4: Modeling of photovoltaic power plants for condition monitoring and power production forecasting





The main objective of this PhD position is to develop advanced models for solar panels and solar energy systems which enable automatic state determination and detection of degrading solar modules. The models will also be used as a basis for prediction of the solar electricity production, especially in a short time frame (hours).


This PhD position will work closely with PhD 3. In this thesis, the development of methodology for state determination based on available, measured parameters will be the main topic. The theme is data analysis and modeling to identify anomalies and prediction of production in the near and distant future.


Description of task


This candidate will access data from selected solar power plants of larger and smaller size, primarily in Norway. Data will be both production data and sensor data where available. Models will be developed, validated and implemented on well instrumented plant. This will be models at the system level. Physical models will be implemented and their ability to detect and identify deviations will be compared with so-called big data approximations based on various forms of pattern recognition. There will be detailed performance and degradation analysis on selected systems and components. The aim is to develop a model based framework that can allow for the determination of performance and detection of deviations in a robust manner for selected technologies. Combined with a sky camera we will also want to try using system models for short term forecasting. This is already a growing field that will become increasingly important in the future. In the long term it is desirable to use the model-based framework as the basis for autonomous analysis of solar power plants' performance, to automatically detect and identify defects and propose measures.




For further information please contact:


Head of department, Dr. Stian Løvold, tlf: +47 941 90 920, e-mail: stian.lovold@its.uio.no
PhD # 1 og # 2: Dr. Jonas Moen, Scientist/Adjunct Associate Professor, jonas.moen@ffi.no,
tlf +47 465 40 561
PhD # 3 og # 4: Dr. Erik Marstein, Scientist/Adjunct Professor, erik.stensrud.marstein@ife.no,
tlf +47 901 17 760

Utlysningen finner du her



Publisert 3. apr. 2017 11:02 - Sist endret 3. apr. 2017 16:15