Large-scale earthmoving activities make up a costly and air-polluting aspect of many construction projects and mining operations, which depend entirely on the use of heavy construction equipment. The long-term jobsites and manufacturing nature of the mining sector has encouraged the application of automated controlling systems, more specifically GPS, to control the earthmoving fleet. Computer vision-based methods are another potential tool to provide real-time information at low-cost and to reduce human error in surface earthmoving sites as relatively clear views can be selected and the equipment offer recognizable targets. Vision-based methods have some advantages over positioning devices as they are not intrusive, provide detailed data about the behaviour of each piece of equipment, and offer reliable documentation for future reviews. This dissertation explains the development of a vision-based system, named server-customer interaction planner (SCIT), to recognize and estimate earth material loading cycles. The SCIT system consists of three main modules: object recognition, tracking, and action recognition. Different object recognition and tracking algorithms were evaluated and modified, and then the ideal methods were used to develop the object recognition and tracking modules. A novel hybrid tracking framework was developed for the SCIT system to track dump trucks in the challenging views found in the loading zones. The object recognition and tracking engines provide spatiotemporal data about the equipment which are then analyzed by the action recognition module to estimate loading cycles. The entire framework was evaluated using videos taken under varying conditions. The results highlight the promising performance of the SCIT system with the hybrid tracking engine, thereby validating the possibility of its practical application.

Large-scale earthmoving activities make up a costly and air-polluting aspect of many construction projects and mining operations, which depend entirely on the use of heavy construction equipment. The long-term jobsites and manufacturing nature of the mining sector has encouraged the application of automated controlling systems, more specifically GPS, to control the earthmoving fleet. Computer vision-based methods are another potential tool to provide real-time information at low-cost and to reduce human error in surface earthmoving sites as relatively clear views can be selected and the equipment offer recognizable targets. Vision-based methods have some advantages over positioning devices as they are not intrusive, provide detailed data about the behaviour of each piece of equipment, and offer reliable documentation for future reviews. This dissertation explains the development of a vision-based system, named server-customer interaction planner (SCIT), to recognize and estimate earth material loading cycles. The SCIT system consists of three main modules: object recognition, tracking, and action recognition. Different object recognition and tracking algorithms were evaluated and modified, and then the ideal methods were used to develop the object recognition and tracking modules. A novel hybrid tracking framework was developed for the SCIT system to track dump trucks in the challenging views found in the loading zones. The object recognition and tracking engines provide spatiotemporal data about the equipment which are then analyzed by the action recognition module to estimate loading cycles. The entire framework was evaluated using videos taken under varying conditions. The results highlight the promising performance of the SCIT system with the hybrid tracking engine, thereby validating the possibility of its practical application.

Submitted by Patrícia Valim Labres de Freitas (patricial) on 2016-03-30T19:29:12Z No. of bitstreams: 1 Gustavo Adami.pdf: 2445130 bytes, checksum: d010b74234328835418d4d0ff2a01f40 (MD5) / Made available in DSpace on 2016-03-30T19:29:12Z (GMT). No. of bitstreams: 1 Gustavo Adami.pdf: 2445130 bytes, checksum: d010b74234328835418d4d0ff2a01f40 (MD5) Previous issue date: 2015-12-23 / Nenhuma / Características existentes no processo de produção da indústria de fluxo contínuo, particularmente a petroquímica, requerem que a medição de eficiência inclua características diferentes da indústria de produção intermitente, tais como a maneira de quantificação dos produtos finais e a natureza das perdas que são consideradas no cálculo da eficiência de produção. Indicadores de eficiência global de produção tipicamente são derivados do OEE (Overall Effective Equipment), proposto por Nakajima (1988) para a indústria de produção intermitente e, por vezes, são utilizados em indústrias de produção contínua sem uma análise prévia de suas limitações. Doze indicadores identificados na literatura foram analisados e comparados com características do processo de produção da indústria petroquímica, obtidos a partir da revisão teórica e de entrevistas com profissionais e pesquisadores dessa indústria. Dessa análise identificou-se que o indicador OAE (Overall Asset Efficiency) apresenta maior aderência em relação à classificação de perdas e às características do processo de manufatura da indústria petroquímica. Os resultados de eficiência global de produção obtidos através da utilização do OAE foram confrontados com os provenientes do OEE e do TEEP (Total Effective Equipment Productivity), com base em dados reais de uma empresa localizada no Pólo Petroquímico do Rio Grande do Sul. Os resultados obtidos através do cálculo de eficiência utilizando o indicador selecionado OAE, se mostraram mais descritivos da realidade da empresa quando comparados com aqueles atualmente utilizados. Outras práticas que geram interferências sobre o cálculo do OEE também foram identificadas nas entrevistas. Ainda, foi identificada a necessidade de uma discussão mais ampla no sentido de melhor definir os conceitos de capacidade e nível de atividade na indústria petroquímica e sua estimação operacional para fins de análise de eficiência, bem como, a incorporação de termos relativos à eficiência de insumos e custeio na análise de eficiência operacional global dessa indústria. / Due to singular characteristics present in the production of continuous flow process industry, especially petrochemical, efficiency measurement require different features of intermittently producing industry, such as how to quantify the final products and the cause of the losses that are considered in the calculation of production efficiency. Production efficiency indicators are typically derived from the OEE (Overall Equipment Effective) proposed by Nakajima (1988) for intermittent production industry, they are sometimes used in continuous manufacturing industries without a prior analysis of their limitations. Twelve indicators identified in the literature were analyzed and compared with features of the petrochemical industry production process, obtained from theoretical review and interviews with professionals and researchers in this industry. This analysis identified that the indicator OAE (Overall Asset Efficiency) has a better production losses structure and fits the characteristics of petrochemical manufacturing process. The results of overall production efficiency obtained using the OAE were compared with results from OEE and TEEP (Total Effective Equipment Productivity), based on manufacturing data from a company located in Rio Grande do Sul petrochemical complex. The results obtained from the efficiency calculation utilizing the selected indicator OAE, are more descriptive of the company situation when compared to those currently used. Other practices that causes interference on the calculation of OEE were also identified in the interviews. It was also identified the necessity for a broader discussion in order to better define the concepts of capacity and activity level in the petrochemical industry and also the operational way define them in order to make analysis of efficiency as well as the incorporation of terms concerning the efficiency of inputs and costing the analysis of overall operational efficiency of this industry.