previous day
all days

View: session overviewtalk overview

08:00-09:00Gather Town open
09:15-10:00 Session K2: Keynote: Olli Ventä: Digitalization of Industry – past, present and future

An overview of digital transformation in heavy industries. A short look back to the development of digital industries is given but the major emphasis is given to the trends, technologies and visions of modern digital technologies relevant to industries. The presentation also brings out the major takeaways from several European research agendas, strategy papers, and white papers – where the author has been either as a leading contributor, a member of a team, or giving comments. 

Olli VENTÄ, D. Sc. Technology, currently a senior citizen, retired from VTT (Technical Research Centre of Finland Ltd.), since 2019. During his career at VTT, Olli Ventä was the leader of the national technology programme Intelligent Automation Systems, in 2001-2004. Thereafter, Dr. Ventä was the leader of several consecutive VTT internal spearhead programmes, all dealing with digitalization of industries. During 2008-14, Olli Ventä was very active in the Finnish Metals and Engineering Competence Cluster Ltd (FIMECC) contributing to the strategies and programme contents in digital manufacturing sector. On EU arenas, Olli Ventä has been active in ECSEL (e.g., an SRA chapter leader of Digital Industry and a LIASE member of Industry4.e - an ECSEL lighthouse), in EFFRA, in BDVA and in SPIRE. Olli Ventä is also an active member of the Finnish Automation Society, currently a Board Member of Finnish Automation Foundation. Dr. Ventä is a Knight, First Class, of the Order of the Lion of Finland, issued by the President of Finland on December 6, 2015.

10:00-10:40Coffee Break / Gather Town
10:40-12:00 Session 4A: Robotics and autonomous systems
Location: Room 1
Quality measurements of trees with a SPAD LIDAR sensor for precision forestry

ABSTRACT. New sensor technologies bring precision forestry closer to reality. Single Photon Avalanche Diode (SPAD) Light Detection And Ranging (LiDAR) has a great potential to outperform conventional linear mode LiDARs in measuring tree parameters in the stand level. So far SPAD technologies have been applied mostly in aerial forestry measurements, but we have applied VTT’s prototype SPAD LiDAR scanner for terrestrial measuring of quality parameters for trees in forest stands. The quality parametes we consider are breast height diameter and trunk bending. In the paper we introduce the prototype SPAD LiDAR sensor and report results for applying it in detecting bends and diameters of trunks in challenging conditions.

Manufacturing Operations as Services based on Robots with Skills

ABSTRACT. Industry 4.0 is transforming the world of manufacturing so that manufacturing-as-a-Service (MaaS) is coming forward via digital manufacturing platforms. Digital manufacturing platforms enable companies to find fast, cost-effective production with others who have manufacturing capacity and as such are increasingly streamlining supply chains. Currently MaaS operators are operating in areas such as machining and 3D printing. The services enabled by digital manufacturing platforms are associated to collecting, storing, processing and delivering data, which describe the manufactured products, the manufacturing processes or other assets that make manufacturing happen. These include material, machines, enterprises, value networks and even factory workers. We take a further step in MaaS by considering robotic manufacturing operations and their usage as flexible service operations. In agile and ultra-flexible operations, where lot sizes go down to one, the setup and execution of new tasks must be instant. We take a data-oriented approach to utilize re-usable robot operations formulated as robot skills, and integrate service requests and system and robot models towards an easily executable manufacturing service system. We present our architectural solutions and show the feasibility of our approach by a simple experimental test with intra-logistics and assembly tasks.

Comparing Performance of Algorithmic and Driver-Planned Routing of Forest Forwarder

ABSTRACT. The time and energy efficiency when collecting felled wood logs from a forest stand are critical for the forestry economics and sustainability. Choice of collection routes varies greatly depending on the experience of the forest forwarder driver. With modern vehicle routing optimization algorithms, this variance could be reduced, and less experienced drivers supported, therefore making the log collection at forest stands more efficient. In this paper the repeated matching (RM) algorithm solving the routing problem for a capacity-limited forwarder is adapted from an earlier work and extended to consider constraints on the number route segments can be driven through and one-way route segments. Furthermore, the forwarder unloading to sorted piles is integrated to the route optimization. The performance of the algorithm is compared to that of a group of forwarder driver students in Finnish national skill competition finals. The case had 164 piles of six timber classes and three depots, some unidirectional segments and some segments with limited number of drives. When our algorithm was run for 8 hours, it performed best amongst the nine finalists, and when it was run for one hour, it performed fourth best.

10:40-12:00 Session 4B: Simulation and digital twins
Location: Room 2
Metallurgical Digital Twin Model for Minerals Processing Plant Optimizers

ABSTRACT. Metallurgical Digital Twin aims for minerals processing site-wide metallurgical and economical optimization of operations. It integrates mining data, consisting of run-of-mine ore characteristics, with a detailed mineral particle-based processing plant model. In addition to the feed ore properties, the Digital Twin model needs to be able to adapt into plant asset availabilities and operating constraints, as well as incorporating the operating costs and metal market price data for producing relevant KPIs for decision making. Adaptation of a processing plant model, based on the feed ore variations and equipment operating parameters and availabilities, implies need for connectivity with mining and real plant data sources. Vice versa, the results of a Metallurgical Digital Twin ‘what-if’ predictions are to be readily used for selecting the best control targets in closed loop manner. This approach for finding the global optimum is referred later on as the Plant Optimizer. Thus, these plant and site level targets generated by the Plant Optimizer are fed into optimizing controls of each process area, which are on top of stabilizing controls and process control systems to implement actual control actions. (Fig. 1). When seeking the global optimum for the plant operation the Plant Optimizer answers questions like: • What are the ore characteristics and chemical recipe? • What is the optimum theoretical economic output of the whole plant? and then sets optimal targets to the Process Optimizers (Model Predictive Controls, MPC). MPC of each process area calculates the optimal control path towards the targets and adjusts respective parameter set points in the process control system for control actions. This paper describes concept and architecture of a Metallurgical Digital Twin that is based on a dynamic mineralogical predictive plant model. The Metallurgical Digital Twin enables predictions of short- and long-term process operating scenarios with alternative control actions and ore characteristics (Fig.2). As an example, a gold processing Digital Twin is presented. The process model is based on first principle equations and chemical reactions with empirical parameters. The flowsheet model was constructed with HSC Chemistry® software, with mineral processing and hydrometallurgy unit models (Fig. 3). In addition to technical aspects of a Metallurgical Digital Twin, some of the key drivers and success factors for a Digital Twin project are presented.

Benefits of simulation assisted automation testing with Apros in Loviisa NPP emergency diesel automation renewal

ABSTRACT. The challenge in nuclear power plant modernization projects is ensuring safe and functional connectivity between existing and new systems. Safety and functionality need to be proven in advance so that the wanted changes are licensed and approved by the operator and the regulator. With the help of dynamic simulation the technical concept can be tested in all design phases and plant operators can be involved early on in the project in functional testing and design work. Dynamic simulation is one part of ensuring smooth commissioning during tightly scheduled nuclear power plant (NPP) outage.

In Fortum's Loviisa NPP the original emergency diesel automation systems are modernized. The role of emergency diesels is to automatically provide power for safety critical components during a total loss of off-site power. The emergency diesel automation consists of safety classified I&C cabinets (analog) and non-safety classified I&C cabinets (PLC) and Human-machine Interphase (HMI) systems (physical controls and touchscreen). The automation functions include different operation and control modes, testing functions and protection functions. Functionalities are too complex to evaluate with pen and paper, especially during fault situations and simultaneous activity of safety classified and non-safety logics. Tight schedule of the NPP overhaul limits the available time for on-site testing.

In this project, the decision of utilizing the Apros simulation software was made during the basic design phase and as a first step, the safety classified systems were modelled. Basic design functionality was tested dynamically with the model. A more detailed Apros simulation model was created based on the function diagrams provided by the automation system supplier, including both safety and non-safety parts. Certain issues were revealed in own testing and during the workshops with automation supplier in both functionality and integration of systems. Using Apros enabled integration of safety and non-safety automation to a plant model which included the emergency diesel generator and its auxiliary equipment. The simulation model was found to be an efficient discussion platform between the supplier and Fortum to find and illustrate the issues. After finishing the detailed design phase, the Apros simulation model was used to carry out functional tests and the simulation results were reviewed together with NPP operators, experts, the automation supplier and also with the Radiation and Nuclear Safety Authority of Finland (STUK).

A simulate HMI was integrated in the Apros simulation model in order to test the design and train the NPP operators. Training sessions were held during development. Changes based on operator feedback from simulator training were implemented in the design. This resulted in a better functionality and a more user friendly HMI. In factory acceptance tests (FAT) the simulation test results were used as validation material and the Apros simulation model was used as a tool for general problem solving. Also during the site acceptance tests (SAT) Apros simulation model was used to study unwanted behavior and to check effects of changing test procedures beforehand.

The functional testing together with simulated HMI created confidence towards functionality and operation of the new automation system. This is an important aspect in NPP organization for executing the project efficiently. Simulation activities were also informed and demonstrated to regulatory body (STUK) and positive feedback was received from them. It was valuable that many improvements were made and issues were found and solved in the early phase of the project, most of them before FAT. Solving these issues during FAT and SAT phase would have been much more costly and time consuming and this way the risk of delaying the whole NPP outage schedule was minimized.

Towards digital twin data portals for existing plant sites based on P&IDs

ABSTRACT. This paper focuses on the challenge of creating digital data portals for existing plants. The method involved the use of the Piping and Instrumentation Diagram (P&ID) as a key document describing the process. The text elements, tag names, of the P&ID components are used as anchor points for providing links to additional static and dynamic plant information. This method can be applied to P&IDs in paper format, given that existing OCR software can recognize the texts. The case study is a part of an Essity paper mill in Nokia, Finland. The P&ID of part of the process was stored as an Autocad P&ID file and it was exported as a PDF document. Prototype software written in Python was able to import that PDF file and create an interactive view that allows the user to click the text tag names of the components and provide links to plant data (as long as mappings exists). Future work will focus on paper P&IDs and explore the potential data sources that can be linked to the P&ID.

10:40-12:00 Session 4C: AI, ML and data based methods
Location: Room 3
Intelligent temporal analysis of coronavirus statistical data

ABSTRACT. The coronavirus COVID-19 is affecting around the world. There are strong differences between countries and regions. People of all ages can be infected but older people and people with pre-existing medical conditions are more vulnerable to becoming severely ill. This research aims to develop unified intelligent temporal analysis methodologies for detecting the fluctuations, trends and severity of the corona situations. Parametric systems are used to adapt the solution for varying operating conditions caused by local areas and groups of people. Recursive updates are used in the parametric models. The analysis is done in a similar way for different subsets. Specific scaling functions can be used in local analysis and for people groups to increase the sensitivity of the temporal analysis. More aggregated material is used for analyzing countries and continents. The norms can be recursively updated and the norm orders related to the scaling functions are updated less frequently.

Data-driven approach to stabilizing an effluent treatment plant (ETP)

ABSTRACT. An industrial manufacturer was having problems with a closed-loop effluent treatment plant (ETP). The unexpected ETP behavior was causing production losses, as the processed wastewater did not meet the quality standards. Process-based COD modelling and Dynamic Centerline Advisor were utilized along with operation analysis in a Data Discovery, which revealed that the ETP operation conditions had not been ideal for biomass growth. Based on the Data Discovery results, the customer revised their water balance management system to compensate for the cyclic nature of their production process. The actions taken were proven to be effective, and the ETP plant is no longer a bottleneck for their production.

Foundations and Case Studies on the Scalable Intelligence in AIoT Domains

ABSTRACT. The Internet-of-Things (IoT) concept is based on networked, mobile, and sensor equipped microelectronic devices. They are capable of reacting to their environment by collecting and processing data, computing, and communicating with other IoT devices and the cloud. The deployment of artificial intelligence (AI) to IoT, referred to as artificial intelligence of things (AIoT), enables intelligent behavior for the whole cyber-physical system (CPS) whether it is designed for human co-operation, completely autonomous operations, or something in between.

The IoT devices, including smart phones and wearables, can be applied in a plethora of applications ranging from building automation and industrial systems to self-driving vehicles and health services. The distributed and growing usage of the connected devices deliver the users more responsive and intelligent support for decision making in a given environment.

The foundation of AI is based on data fed to algorithms for machine learning (ML). They require a lot of processing power due to the amount of data and recursive/concurrent nature of calculation. Until recently this has been accomplished mainly in the cloud environment, where the raw data is uploaded into. This exposes all the data, even private and sensitive data, to the transmission phase and processing system. In conjunction with IoT there is a possibility to perform ML closer to the origin of data concerning local intelligence. It means that only the results of local or edge ML are transmitted to cloud for more general aggregation of AI. Local systems do not need to send the raw data anymore, which helps on prevailing the privacy and security of the data. This type of ML is referred to as federated/collaborative learning (FL).

This study focuses on finding the existing and/or recommended solutions for up-to-date AI close to the devices. At first, the definitions of devices are reviewed in order to find out classifications of their capacity to contribute for the computation and scalability. Secondly, the other computing and serving options between the devices and the cloud are studied. Those are referred to as Fog/Edge services, and they are more stationary than the IoT devices. Thirdly, the facts learned are being applied in two use cases in order to support the discussion and applicability of AIoT in practice.

The main conclusion is that there are no silver bullets for solving all the requirements. Instead there are multiple options from mutually connected devices via middle layer support to cloud services, and distributed learning, respectively.

Machine Learning-based Classifier in the Analysis of Nuclear Power-specific Requirements

ABSTRACT. Typical nuclear power plant projects include the sheer volume of descriptive and non-harmonized requirements, which have to be managed and fulfilled. These requirements are typically hard to interpret and humans' very limited ability to concentrate on a specific task together with a large number of requirements usually cause main errors in the analysis of the requirements. By utilizing artificial intelligence in the analysis of nuclear power plant requirements, designers' decision-making in classification and allocation of requirements could be facilitated and thus, errors reduced.

Fortum developed a machine learning-based requirements classifier utilizing state-of-the-art natural language processing (NLP) and integrated it with a requirements management system. The classifier categorizes nuclear power industry-specific requirements into pre-defined categories. The categories have been determined based on processes and design disciplines of Fortum's nuclear engineering departments that are responsible for fulfilling requirements.

The very promising results include predetermined requirement classes, manually gathered and labeled data, comparison of three models and their classification accuracies. In addition to these results, the established classifier was integrated with the requirements management system. Future development suggestions include focusing on atomizing (i.e., splitting up) long, especially multiclass requirements, combining similar ones and checking requirements syntax based on suggestions generated by an AI-model. Furthermore, new and practical requirement classes and hierarchies are suggested to be developed while also improving current datasets both quantitatively and qualitatively.

12:00-13:00Lunch Break / Gather Town
13:00-14:20 Session 5A: Control, modelling
Location: Room 1
Learning-based approach for adaptive MPC tuning

ABSTRACT. Automated process control has an important role in operating the production processes in resource efficient manner. However, the process nonlinearities, model-plant mismatch, and conservative tuning in Model Predictive Control (MPC) prevents to achieve its full potential. Learning-based methods can provide impacts on resource-efficient production by improved process control. One option is to use model-free, algorithmic solutions to adaptive on-line tuning of the MPC. The approach used in this study is based on information theory and it is demonstrated with a benchmark simulator. The results indicate that process performance benefits on adaptive MPC weighting in terms of improved control performance and decreased actuating energy.

Model-based on-board post injection control development for marine diesel engine

ABSTRACT. The increasing demands for reducing fuel con- sumption and emissions in contemporary technology so- lutions lead to the use of more sensors, actuators, and control applications. With this increasing engine com- plexity, the feedback design is complex due to the cou- pling between inputs and combustion parameters. To be able to design the controller systematically, model predictive control (MPC) comes to the scope because of its advantages in the design of multi-input multi- output (MIMO) systems, especially with its constraints handling ability and performance in simultaneously op- timizing the engine fuel efficiency and emission reduc- tion. Multi-injection is one of the promising techniques for achieving better engine performance. In this work, post injection control is implemented utilizing MPC MIMO strategy with the target of exploring the pos- sibility of reducing emissions and improving engine ef- ficiency by controlling post injection duration and in- jection timing. The workflow of the MPC controller de- sign, from control oriented model (COM) establishing, to MPC problem formation and solution methodology are discussed in this work. Moreover the embedded im- plementation of the MPC controller purely by Matlab Simulink for rapid control prototyping design is con- ducted. The simulation result demonstrated the ability of the controller’s tracking performance and shown con- stitute the preliminary step towards the nonlinear com- bustion model based multi-injection MPC design. The systematic model-based controller framework developed in this work can be applied to other control applications and enables a fast path from design to test.

Model predictive control in multiple injection strategy for maritime diesel engines

ABSTRACT. Lowering the emissions excessively under the legislation limits would increase the engine operating costs such as higher fuel consumption. The challenge is to fulfill the emission standards and to maximize the engine efficiency. Overcoming this problem requires more advanced combustion control than the traditional engine tuning and calibration. There have been several new approaches in the engine combustion control, which are based on new low temperature combustion technologies, e.g.\ RCCI (reactivity controlled compression ignition), HCCI (homogeneous charge compression ignition) and PPC (partially premixed combustion). Control in these is complicated however, and they are based on controlling the cylinder pressure by several successive fuel injections during each engine cycle. Engine efficiency can be increased by maintaining the cylinder pressure at its highest level throughout the combustion. Multiple injection strategy is one of the new methods being used to maximize the cylinder pressure. It is an alternative approach to single injection combustion, which has been proved to be more efficient in terms of noise and emission reduction, and fuel consumption. The multiple injection enables better control of the fuel distribution. In this work, a model predictive control (MPC) is applied for multiple injection strategy in a maritime diesel engine. The aim is to predict and control the engine combustion parameters cycle-to-cycle.

13:00-14:20 Session 5B: Energy and electrification
Location: Room 2
Electrification as a Solution to Carbon Neutral Society

ABSTRACT. This study demonstrates how various power system engaged part-solutions can be combined as a functional, operable, and climate neutral electric power system. This is a simulation study taking the Finnish electric energy system as a case study. A chief aim here is to study how investments e.g. in wind and solar power production, heat pumps on a large scale, and increased number of chargeable electric vehicles influences the other components and operability of the existing power system. In this way we can obtain new insights on how different part-solutions function together and how they imply on energy prices and emissions. This is a simulation study made by freely available IRENA FlexTool modelling tool. The modelling results are combined with an analysis of what policy measures and instruments the transition to modelled 2030 and 2050 scenarios are likely to require to become actually implemented in the society. In other words, what kind of political regulation is needed to achieve the goals of functional and climate neutral power system. It is very important to analyse this problem in a systemic perspective, how different decisions and actions effect on the system level, instead of typical local optimization ignoring the systemic effects. In this study we will present different scenarios about the structures and functionality of the national power system on 2030 and 2050. According to these results we will assess the feasibility of different options, and what kind of regulation will be needed to achieve this desired state. This study is a part of the research project “Transition to a resource efficient and climate neutral electricity system” (EL-TRAN), financed by The Strategic Research Council of Academy of Finland.

A Novel Hybrid Energy Production Module for securing Energy Availability in Rural Areas

ABSTRACT. Background Reducing the carbon footprint and using renewable energy are key factors in the implementation of both national and global climate strategies. The Finnish National Energy and Climate Strategy emphasizes the use of renewable energy, decentralized production of power and heat, and two-way energy markets. Decentralized energy production with multiple energy sources enables self-sufficiency also in sparsely populated areas, as well as in situations where energy demand has temporarily increased, e.g. due to temporary utilization or during storms. The consequences of climate risks can be managed more efficiently, economically and environmental-friendly with local combined heat and power solutions.

Aim In rural areas, energy self-sufficiency plays a key role in ensuring vital operations for both livestock and living. Energy outages are typically caused by environmental disasters such as snowstorms and may last for several days. To overcome this challenge, self-sufficient energy production systems are seen necessity. The challenge is not a new, but the practical implementation has typically been based on oil or biomass energy sources, without optimizing the use of raw materials or emissions. To meet this challenge, this paper introduces a framework for a demand-based mobile hybrid energy production module with the capability to produce both thermal and electrical energy at anytime and anywhere, more efficiently and sustainable manner. The combination of solar, biomaterial-based combustion process, thermoelectric generators and energy storages with backbone of AI assisted optimization enable cost and energy efficient, and sustainable demand-driven energy production.

Materials and methods The hybrid module is designed to operate on its own, in rural, sparsely populated areas, despite of climate disasters, and other unexpected outages. It integrates the generation and storage of thermal and electric energy, with maximum capacity targets of 60kWth and 6kWe, guaranteeing energy availability for basic functions in living and farming context. Predictive and optimizing system is based on open data analysis on local weather conditions and raw material prizes combined with statistical analysis and algorithm development. Further, the combustion process enables the use of variable bio-based waste material, such as wood chips and bio waste and manure. This paper introduces the developed hybrid module, justifies the chosen technology solutions, first results on practical implementation and test runs. Further, methodology to verify the efficiency and variable bio-based raw materials by Design of Experiment (DoE) with Box-Behnken -methodology is utilized to define relevant process parameters and test series for the optimization assessment and minimizing pollutant emissions.

Results Hybrid module is a technology solution enabling micro scale energy production anywhere, and environmentally friendly manner. The module is mobile, size of a small sea container. Module’s intelligent analytics utilizes solar energy and energy storages when appropriate and utilizes combustion process to fulfill the gap between energy demand and production. Control is web based, enabling operation at distance. The process status and up to date measurement values can be reviewed online: energy storages is optimized based on long term weather forecasting and consumption profile. Table 1 introduces the equipment to produce energy in multitude manner. The layout, installation and measuring parameters are further presented in the full paper. The experimentation is conducted in two phases after initial setups. First, the process initialization is conducted with preparatory measurements, and then the raw material tests runs are conducted based on Box-Behnken DoE design, variables introduced in Table2. The first measurement results are presented if Fig.1 The figures show clear dependence on process parameters and measured emissions over time. Emission measurements were taken with calibrated Testo380 meter. The major emission caused by the process was CO, while other emissions remained low. According to the time behavior presented Fig.1, the most unstable operating period is in the beginning of process, where the temperature increases rapidly, and most emissions rises remarkably. The full paper will present a comprehensive analysis over measured and process parameters in thermal energy production in regard of bio combustion process and emission measurements. Further, results are analyzed over produced heat. Conclusion The added value of this research is achieved by optimizing self-sufficient energy production based on demand response, local weather conditions, national electric stock market prize and raw material costs. Optimization of combustion process minimizes the GHG emissions caused by the utilization of bio-based waste and raw materials. The existence of the designed hybrid module is advocated by optimization of both economic and environmental benefits. To maximize the beneficiaries, the optimization algorithms can be applied also to traditional combustion processes.

Simulation tool to analyze the productivity and energy consumption of electric mining vehicles

ABSTRACT. The transfer from conventional mining methods toward digitalization, automation and electrification is going on worldwide. The mining process efficiency needs to be improved in the future, because the mines are in more difficult places, far away from the infrastructure and very deep underground hence the high profitability is more difficult to achieve. Therefore, the mining process needs to be more energy efficient in the future to keep the business profitable. Accurate information about the productivity and process optimization is needed to achieve suitable decisions. The operation of mine needs to be modeled to find the most critical aspects in reliability and productivity point of view. In this study, an electric grid and mining vehicle traffic simulator, called MineGame, is created to model, visualize and study fleet-level operation of mine with simplified physics. Open source coding software is used to create easy-to-use tool to support complex decision making. MineGame enables the analyze of productivity and energy consumption of varying mining machine fleets, mine layouts, task management decisions and traffic rules. Engineers, who design mine operation, sale persons of mining devices and trainers may use MineGame.

14:20-15:00Coffee Break / Gather Town
15:00-16:15 Session C2: AutomaatioStudio panel discussion and closing

Tapio Heikkilä, Tuula Ruokonen, Sonja Mäkinen, Juha Hirvonen, Antti Jaatinen, Mats Friman, Helena Leppäkoski, Tero Hietanen, Jouni Aro

16:15-18:00Gather Town open