Identification and predictive capabilities of IBM Watson in industrial wastewater treatment
Diplomityö
2018
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2018053125045
https://urn.fi/URN:NBN:fi-fe2018053125045
Tiivistelmä
With current rate of development of technologies and researches in different spheres it may be challenging for scientists to be kept up-to-date on. One of the main reasons is that only small part of this information is suitable for coherent analysis, integration and classification for end users. The issue lies in deployment and extraction of data from many sources and conversion into suitable formats.
For efficient wastewater treatment it is absolutely crucial to achieve precise control and management of operating parameters. Temperature, flowrates, concentration of inlet compounds and amount of added chemicals may cause significant fluctuations of the wastewater purification efficiency. Some of the impurities promote the growth of the hazardous bacteria and pathogens, which are the “top priority” for removing during the wastewater treatment processes. In spite of the fact that all purification processes are tightly-controlled, the concentration of inlet compounds may fluctuate heavily and does not correlate with processing parameters, and as a result, it may affect the treatment efficiency.
New technologies for efficient data analysis allow to process and evaluate big datasets according to user’s request. These systems are able to understand and analyze different types of data, such as laboratory results, industry and technical-specific content (chemistry, pharmaceuticals, medicine, economy), compare the scientific results, summarize numeric values through the instrumentality of predictive modeling and statistical algorithms.
One of the potential solution is IBM Watson - a question-answering system, which uses combination different predictive algorithms. In question-answering system the question may be defined as an initial input in data processing. Program extracts key elements of the questions and provides precise answers according to the user’s input. IBM Watson Analytics is a cloud-based system, which is used for data analysis and pattern recognition. This tool is suitable for patterns and dependencies searching in data results and determination of the statistical drivers of analyzed values. Application of IBM Watson Analytics for analysis of wastewater purification process related data provides an opportunity to effectively investigate and evaluate operating parameters thus enabling in-depth analysis of the process. Manual estimation of the most significant parameters is time-consuming and, in some cases, impossible due to the high amount of data and large quantity of variables. Also, non-controlled inlet parameters (such as a concentration of impurities in wastewater) may be taken into account as affecting factors for purification processes.
For efficient wastewater treatment it is absolutely crucial to achieve precise control and management of operating parameters. Temperature, flowrates, concentration of inlet compounds and amount of added chemicals may cause significant fluctuations of the wastewater purification efficiency. Some of the impurities promote the growth of the hazardous bacteria and pathogens, which are the “top priority” for removing during the wastewater treatment processes. In spite of the fact that all purification processes are tightly-controlled, the concentration of inlet compounds may fluctuate heavily and does not correlate with processing parameters, and as a result, it may affect the treatment efficiency.
New technologies for efficient data analysis allow to process and evaluate big datasets according to user’s request. These systems are able to understand and analyze different types of data, such as laboratory results, industry and technical-specific content (chemistry, pharmaceuticals, medicine, economy), compare the scientific results, summarize numeric values through the instrumentality of predictive modeling and statistical algorithms.
One of the potential solution is IBM Watson - a question-answering system, which uses combination different predictive algorithms. In question-answering system the question may be defined as an initial input in data processing. Program extracts key elements of the questions and provides precise answers according to the user’s input. IBM Watson Analytics is a cloud-based system, which is used for data analysis and pattern recognition. This tool is suitable for patterns and dependencies searching in data results and determination of the statistical drivers of analyzed values. Application of IBM Watson Analytics for analysis of wastewater purification process related data provides an opportunity to effectively investigate and evaluate operating parameters thus enabling in-depth analysis of the process. Manual estimation of the most significant parameters is time-consuming and, in some cases, impossible due to the high amount of data and large quantity of variables. Also, non-controlled inlet parameters (such as a concentration of impurities in wastewater) may be taken into account as affecting factors for purification processes.