Welcome to "Enlitify", a captivating series that delves into the academic journeys of our esteemed Enlighters. As a company deeply rooted in knowledge and innovation, Enlitia takes pride in supporting its team members in their pursuit of excellence, both professionally and personally. "Enlitify" draws its inspiration from "Edify," reflecting our commitment to intellectually and morally uplift our Enlighters.
This series consists of interviews made with our amazing Enlighters in order to share their experience, the struggles and the upsides.
I took a master's degree in electrical and computer engineering, specializing in the energy domain. This academic path mostly aligns with the specific knowledge needed to work on new algorithms or improve existing ones for the energy industry (renewables, grid, ...), which are skills appreciated for analytics, data science and machine learning at Enlitia.
My thesis title is “Cellular Time Activation Networks, a novel approach applied to photovoltaic anomaly detection”. It proposes a new conceptualisation for data coherence in dynamical systems, motivated both by the desire to experiment a novel idea for representing large interconnected systems, and the need to detect faults in industrial-scale photovoltaic farms.
Given the self-imposed constraints of maintaining data privacy and being system-agnostic in my approach’s design, the conclusion that it is possible to find (in)coherence using only the temporal behaviour of networked elements, without even knowing what they represent, contributes to all the scientific fields that study dynamical networks with private subnetworks.
My tool could identify faults in photovoltaic systems on neighbouring inverters without directly sharing historical data or centralization, synchronization, normalization, or outlier cleaning. This outcome opens the possibility of using different stakeholders’ data to benefit each other in finding anomalies and inconsistencies between their systems without heavy data gathering, processing steps or privacy concerns.
Since the thesis’ application was photovoltaic systems, using actual data from one of Enlitia’s clients, the expertise gained perfectly aligns with my work. The thesis’ conclusions showcase tangible results for a client, which is relevant in determining the viability of providing my tool’s algorithm as a service.
Firstly, Enlitia taught me how to write proper software with the help of its senior developers. This technical improvement was crucial for the implementation phase of the thesis, which went smoothly and resulted in an almost complete product ready to deploy for clients. Enlitia also kindly leased the data used for testing, with the condition of not revealing any client details. Furthermore, the work flexibility was also a plus, allowing me to prioritize the academic work whenever necessary.
This was indeed challenging. During the whole thesis duration, I was already a full-time employee, which came with the challenge of juggling academic work and my position’s work. My strategy was to try and be completely transparent with my managers and consider my academic tasks part of the “sprint planning” so that my time commitment to each side was accounted for and planned. Another thing that helped me was always trying to start the day with academic work to avoid delaying it until I was too tired to think or code.
I plan to continue working at Enlitia, since it aligns with my desire of continuous innovation and contributing to the energy industry for a more sustainable society. The future still needs to be determined, academically speaking. I still plan to contribute scientific work to the community through publications, and there's always the possibility of going back and enrolling in a Ph.D., given the opportunity. For now, I will focus on educating myself on other skills the academic world didn't provide.
Meet David, a 22-year-old visionary born in Madeira, Portugal, who in 2018 went to Porto to pursue a master's degree in electrical and computer engineering, specialising in the energy domain.
His ambition and enthusiasm drive his exceptional problem-solving skills, making him a dedicated student and worker. David passionately commits to forging innovative solutions for the future of energy systems, excelling in automation and machine learning.
You can connect with David on LinkedIn to explore his journey and engage in the discourse of pioneering advancements in the field.
You can access his complete thesis named "Cellular Time Activation Networks, a novel approach applied to photovoltaic anomaly detection." here.