Loading...
Publication
Melhoria de processos internos da equipa de assistência técnica e serviços técnicos
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 5.28 MB | Adobe PDF |
Abstract(s)
Este trabalho visou otimizar os processos do departamento técnico da France AIR Portugal, empresa especializada em AVAC, através da aplicação da metodologia Action Research. Inicialmente, foram diagnosticadas ineficiências na gestão de recursos, orçamentação e comunicação interna. Na fase de desenvolvimento, concebeu-se e integrou-se um portal de consulta de obras em lista de espera, um sistema automático de consulta de acessórios para orçamentação e um aperfeiçoamento do fluxo de informação para dar entrada das encomendas de serviços. Paralelamente, implementaram-se modelos de previsão recorrendo ao python, destacando-se os modelos Holt-Winters, XGBoost e ARMAX, tendo sido feita a otimização de hiperparâmetros via GridSearchCV para os modelos machine learning, culminando na criação de um dashboard iterativo. As fases de verificação e refinamento envolveram o ajuste contínuo das ferramentas em conjunto com a equipa e aos utilizadores finais. Os resultados revelaram diminuição substancial da repetição de tarefas, aumento significativo do ritmo de emissão de orçamentos e melhoria na fiabilidade das previsões, permitindo uma alocação de recursos mais eficiente e redução de custos operacionais. A monitorização final definirá indicadores de desempenho para garantir a sustentabilidade das melhorias.
This dissertation aimed to optimize the processes of the technical department at France AIR Portugal, an HVAC specialist, through the application of the Action Research methodology. Initially, inefficiencies were diagnosed with resources management, quotation workflows and internal communication. During development, pendant projects could be consulted through a portal which was designed and integrated, alongside an automated accessory lookup system for quotations and an enhanced information flow for processing service orders. In parallel, forecasting models were implemented in Python, most notably Holt-Winters, XGBoost and ARMAX, with hyperparameter defined automatically via GridSearchCV for the machine learning models, culminating in the creation of a dynamic and interactive dashboard. The verification and refinement phases involved continuous adjustment of these tools in collaboration with the team and end users. Results revealed a substantial reduction in task repetition, a significant increase in quotation output and improved forecasting reliability, enabling more efficient resource allocation and lower operational costs. Final monitoring will define performance indicators to ensure the long-term sustainability of these improvements.
This dissertation aimed to optimize the processes of the technical department at France AIR Portugal, an HVAC specialist, through the application of the Action Research methodology. Initially, inefficiencies were diagnosed with resources management, quotation workflows and internal communication. During development, pendant projects could be consulted through a portal which was designed and integrated, alongside an automated accessory lookup system for quotations and an enhanced information flow for processing service orders. In parallel, forecasting models were implemented in Python, most notably Holt-Winters, XGBoost and ARMAX, with hyperparameter defined automatically via GridSearchCV for the machine learning models, culminating in the creation of a dynamic and interactive dashboard. The verification and refinement phases involved continuous adjustment of these tools in collaboration with the team and end users. Results revealed a substantial reduction in task repetition, a significant increase in quotation output and improved forecasting reliability, enabling more efficient resource allocation and lower operational costs. Final monitoring will define performance indicators to ensure the long-term sustainability of these improvements.
Description
Keywords
Optimization process improvement efficiency forecasting methods python HVAC Melhoria contínua Eficiência Métodos de previsão Dashboard AVAC Otimização