Silmäile
Hakuohje

Osuva

Osuva on Vaasan yliopiston avoin julkaisuarkisto. Osuva sisältää Vaasan yliopiston omat julkaisut, opinnäytteet ja tieteellisten artikkeleiden rinnakkaistallenteet. Osuvaan sisältyy julkaisujen viitetietoja, tiivistelmiä ja kokotekstejä. Sähköisten arkistokokoelmien sisältö ei ole luettavissa verkossa.

Kokoelmat Osuvassa

Viimeksi tallennetut

  • ENERGY CONSUMPTION PREDICTION FOR ROBOTIC TASK EXECUTION USING MACHINE LEARNING AND HIGH-LEVEL OPERATIONAL DATASETS
    Abuzar, Muhammad (2026-05-25)
    Pro gradu -tutkielma
    Abstract This study examines whether energy consumption can be predicted using machine learning methods on high-level operation data sets to execute robotic tasks. With the growing integration of robotic systems in industrial and service settings, there is a strong need to enhance their energy efficiency to cut down on operation costs and to promote the cause of sustainable development. A quantitative approach was used on a designed dataset of 500 observations of robotic tasks and included processing time, task, sensor, environmental and operational status indicator features. Four predictive models were created and assessed a mean-based baseline model, a linear regression model and a random forest regressor. Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) were used to evaluate model performance on a test set held out. These findings show that the three models had widely similar predictive accuracy, with the lowest MAE (0.5385 kWh) and the lowest RMSE (0.6152 kWh) values of the random forest and the baseline models respectively. Feature importance analysis of the random forest model assigned the highest variance reduction scores to processing time and accuracy, at approximately 0.284 and 0.272 respectively. These scores reflect an algorithmic bias of the random forest toward continuous variables, which accumulate importance through a greater number of candidates split points, rather than confirmed causal influence on energy consumption. All other features contributed at much lower levels. The results indicate that machine learning models do not make significant improvements compared to a basic baseline prediction when they are trained using high-level operational data. This result implies that the dataset has not enough informative predictors to enable precise energy modeling, probably due to the failure to capture the low-level physical processes that directly determine energy consumption in robotic systems with high-level task descriptors. The paper presents a practical and replicable assessment system of data-driven energy forecasting in robotics, the main features of the datasets that constrain the performance of machine learning, and the significance of more detailed, richer data to achieve successful energy prediction in robotic tasks.
  • THE IMPACT OF ENERGY EFFICIENCY INITIATIVES ON OPERATIONAL PERFORMANCE IN NORDIC AUTOMOTIVE MANUFACTURING FIRMS
    Haider, Waseem (2026-05-26)
    Pro gradu -tutkielma
    This research investigates how the energy efficiency projects have affected the operational performance of Nordic automotive manufacturing companies, in this case, Volvo Group and Scania, in 2013-2023. The study examines the effect of the degree of energy intensity, investment in energy efficiency, and use of renewable energy on productivity and efficiency in operations. Quantitative longitudinal research design (secondary data analysis) was used to conduct a study. Annual reports, sustainability reports, and international data sources such as Eurostat and the International Energy Agency were used to gather data. Correlation analysis, regression analysis, and descriptive analysis were some of the statistical methods that were used to investigate relationships between variables. These results reveal that energy efficiency programs have a positive impact on improving productivity, reducing costs, and achieving sustainable operation. The research also sheds light on environmental regulation and technological innovation as important measures towards ensuring the generation of energy-efficient manufacturing systems. Overall, the study proves that energy efficiency can be a long-term capability that can be used to promote competitive advantage and sustainability. Looking at the results: energy efficiency investment was the most powerful motivator of operational results positively, with 97% of productivity differences attributable to energy saving initiatives for the 3 firms each firm used a different set of parameters to reach good results. And, future studies need to be extended to other larger geographic settings, more sampled companies and the increasing importance of electrification and supplies chain energy dynamics to further develop the arguments revealed by this research. Keywords: Energy saving, functional performance, automobile production, sustainability, renewable energy, Nordic companies, productivity
  • A Comparative Analysis of Statistical and Machine Learning Methods for Retail Demand Forecasting to Support Operational Planning
    K C, Binita; Dangol, Rajib (2026-05-25)
    Diplomityö
    Retail demand forecasting is essential for businesses in their operational planning and supply chain management. Since the rise in the level of competitiveness in the markets and the adoption of data-driven decisions is high, there is a growing need for forecasting in businesses. For retailers, forecasting is essential because it aids in planning the workforce, distribution, reorder levels, and inventories. In contrast to the conventional statistical approaches that have dominated forecasting, the machine learning algorithms have introduced alternative models for complex demands. There is, however, little certainty about the relative effectiveness and viability of these forecasting approaches in the retail environment. This study aims to compare statistical and machine learning techniques to retail demand forecasting and the relevance of the approaches in retail operational planning. The research will compare the accuracy of the forecasted values, the behaviour of the models, and how helpful the approach is to the decision-making process in the retail sector. This research work has been conducted based on quantitative and empirical techniques. The empirical analysis has been conducted based on publicly available sales transaction data obtained from the UCI Machine Learning Repository. The dataset has been pre-processed through data cleaning, data aggregation, outlier handling, and feature engineering processes. In addition to that, the fixed-origin-hold-out validation technique has been used for model comparison for the Naive, ETS, SARIMA, XGBoost, Random Forest and Neural Network models. Forecasting accuracy has been calculated using forecasting accuracy metrics MAE, RMSE, and WMAPE. The findings from the analysis prove that machine learning approaches outperform conventional statistical methods in managing the data with high volatility in the retail industry. From the results obtained, Random Forest emerged to be the most reliable method in generating unbiased predictions as it performs well in balancing bias and variance. The study shows that precise forecast results could improve operational planning through reduced risks while decision-making on inventory management. The research is an addition to the existing forecasting and retail analytics body of knowledge through employing a holistic analysis approach using public retail datasets to benchmark various forecast methods. Future studies are recommended to incorporate external exogenous factors in the prediction model along with a hybrid architecture system.
  • Managing Data Interoperability Risks in Digital Twin-Enabled Battery Management System (BMS) Projects in Smart Manufacturing Industry
    Islam, Shoumik (2026-05-15)
    Diplomityö
    This thesis investigates how complex projects such as Digital Twin (DT) enabled smart manu- facturing industry based projects can face data interoperability failures and how those can be identified as project-level risks and managed proactively. The research foundations is based upon development of a Electric Vehicles (EV) Battery Health Management System (BHMS), where multiple engineering disciplines team exchange data across organizations having in- compatible tools, standards, and organizational boundaries. Existing industry frameworks are exploring majorly “technical risk” aspects, however the project risk management aspect has been underexplored that classifies such failures under a generic label, obscuring the distinct root causes that require fundamentally different responses. A qualitative, literature-driven conceptual case study combined with innovative research based design science principles was adopted to analyze the delivery context of this project. Classic Project management tool has been explored and modified within the existing project risk management framework such as Work Breakdown Structure (WBS), Design Structure Matrix (DSM), and Risk Breakdown Struc- ture (RBS). Those tools were experimented against a six-level interoperability taxonomy to de- compose and figure out where and how reliability breaks down. Six distinct interoperability failure types were identified at specific stakeholder boundaries. The DSM analysis confirmed that every rework cycle in the BHMS case crosses at least one organizational boundary which establishes the fact that it is technically challenging to resolve interoperability failures indepen- dently as a team. These structural findings were later used to develop a three-phase Project Digital Twin (PDT) framework namely pre-deployment risk structuring, execution-phase mon- itoring through five project management based observable indicators, and level-specific re- sponse activation that replaces ad hoc investigation with targeted corrective logic. The frame- work is conceptual and has not been validated in a live project environment, which defines the scope for future empirical testing. The findings provide a framework establishing the fact that, interoperability risk can be made visible and required actions can be modified and adopted using standard project management instruments without requiring specialized data engineer- ing expertise. This creates a foundation for more proactive, structured risk control in complex DT-enabled manufacturing projects.
  • Ultra-Short-Term Solar PV Power Forecasting Using Gradient Boosting Trees
    Ben, Simo (2026-05-25)
    Diplomityö
    As solar PV power generation becomes more significant in the Nordic power systems, opera-tion, flexibility management and local energy management are also becoming more challeng-ing. In high-latitude regions, such as Helsinki, Finland, cloud-induced irradiance fluctuations, large sea-sonal variations in Solar elevation, shorter daylight in winter and snowy/icy surface conditions can greatly influence PV production, making power forecasting more challenging. These conditions lead to forecast uncertainty at very short time horizons and demand fore-casting techniques with the ability to forecast both fast power variations and seasonal operat-ing patterns. The aim of this thesis is to examine the possibilities of ultra short-term (UST) PV power forecasting with a 10–60-minute forecast horizon for the Helsinki Kumpula measure-ment site data. A supervised machine-learning approach using Gradient Boosting Regression Trees (GBRT) is designed to model the non-linear relationship between PV power output and environmental predictors. The input features are lagged PV power measurements, irradiance-related features, cloud cover, air temperature, wind speed, relative humidity, solar-geometry features, clear-sky normalization, and snow-sensitive predictors. These features are intended to depict the most recent system behavior, evolving me-teorological conditions, and seasonal effects in the high latitudes. A time-series-aware validation approach is used to evaluate model performance, and a persistence benchmark is used for com-parison. RMSE, NRMSE, MAE, R2 and forecast skills are used to measure forecast accuracy. The outcomes show that the model results in the best accuracy for the smallest forecast time used in the experiment, 10 minutes, with an RMSE of 1.94 kW and an nRMSE of 10.03%. The RMSE and nRMSE values are also slowly increasing as the prediction horizon increases, increasing to 2.41kW and 12.46% re-spectively after 60 minutes. The GBRT model is superior to the persistence in fore-casting skills as seen from the skill scores of 0.22 at 10 minutes and 0.27 at 60 minutes, which are positive values. The results also indicate that recent PV power observations are more significant than the other predictors for short forecast leads, while the meteorological, solar-geometry, and snow-related predictors gain in significance with increasing forecast leads. The central idea of this thesis is a systematic and transparent forecasting framework based on GBRT for the ultra-short-term prediction of PV power in Nordic high latitude conditions. The proposed solution enhances the short-term situational awareness of PV generation, thus paving the way for the seamless inte-gration of solar power into flexible power systems. The results also underpin future hybrid ap-proaches to forecasting, based on machine learning and physical irradiance modelling and sky-camera-based cloud information.