Applied deep learning and data science with a human-centric and data-centric approach

The term Artificial intelligence acquired a lot of baggage since its introduction and in its current incarnation is synonymous with Deep Learning. The sudden availability of data and computing resources has opened the gates to myriads of applications. Not all are created equal though, and problems might arise especially for fields not closely related to the tasks that pertain tech companies that spearheaded DL. The perspective of practitioners seems to be changing, however. Human-Centric AI emerged in the last few years as a new way of thinking DL and AI applications from the ground up, with a special attention at their relationship with humans. The goal is designing a system that can gracefully integrate in already established workflows, as in many real-world scenarios AI may not be good enough to completely replace its humans. Often this replacement may even be unneeded or undesirable. Another important perspective comes from, Andrew Ng, a DL pioneer, who recently started shifting the focus of development from “better models” towards better, and smaller, data. He defined his approach Data-Centric AI. Without downplaying the importance of pushing the state of the art in DL, we must recognize that if the goal is creating a tool for humans to use, more raw performance may not align with more utility for the final user. A Human-Centric approach is compatible with a Data-Centric one, and we find that the two overlap nicely when human expertise is used as the driving force behind data quality. This thesis documents a series of case-studies where these approaches were employed, to different extents, to guide the design and implementation of intelligent systems. We found human expertise proved crucial in improving datasets and models. The last chapter includes a slight deviation, with studies on the pandemic, still preserving the human and data centric perspective.

High-strength steels manufactured via Laser Powder Bed Fusion: effect of post-process heat treatments and surface finishing on microstructure and mechanical properties

Laser Powder Bed Fusion (LPBF) permits the manufacturing of parts with optimized geometry, enabling lightweight design of mechanical components in aerospace and automotive and the production of tools with conformal cooling channels. In order to produce parts with high strength-to-weight ratio, high-strength steels are required. To date, the most diffused high-strength steels for LPBF are hot-work tool steels, maraging and precipitation-hardening stainless steels, featuring different composition, feasibility and properties. Moreover, LPBF parts usually require a proper heat treatment and surface finishing, to develop the desired properties and reduce the high roughness resulting from LPBF. The present PhD thesis investigates the effect of different heat treatments and surface finishing on the microstructure and mechanical properties of a hot-work tool steel and a precipitation-hardening stainless steel manufactured via LPBF. The bibliographic section focuses on the main aspects of LPBF, hot-work tool steels and precipitation-hardening stainless steels. The experimental section is divided in two parts. Part A addresses the effect of different heat treatments and surface finishing on the microstructure, hardness, tensile and fatigue behaviour of a LPBF manufactured hot-work tool steel, to evaluate its feasibility for automotive and racing components. Results indicated the possibility to achieve high hardness and strength, comparable to the conventionally produced steel, but a great sensitivity of fatigue strength on defects and surface roughness resulting from LPBF. Part B investigates the effect of different heat treatments on the microstructure, hardness, tensile and notch-impact behaviour of a LPBF produced precipitation-hardening stainless steel, to assess its feasibility for tooling applications. Results indicated the possibility to achieve high hardness and strength also through a simple Direct Aging, enabling heat treatment simplification by exploiting the microstructural features resulting from LPBF.