Data, Algorithms, Sound and Design (Research)

Data, Algorithms, Sound and Design (Research)
Raytheon Amphitheater
9 - 11 AM

Presentations by Paolo Ciuccarelli, Fábio Duarte, Sara Lenzi, Simone Mora, Israel Salmon Ruiz and Aiur Retegi Uria.

To attend virtually, register here.

As an intentional, design-driven approach to data sonification continues to evolve at the intersection of scientific analysis and artistic exploration, this session highlights emerging research directions and community-driven initiatives, with topics ranging from autographic sonification to the application of machine learning in machine listening. Updates from the field include the Data Sonification Archive, now featuring over 500 projects; the newly launched Data Sonification Awards, which received 80+ applications and engaged 15 jurors; and the Dagstuhl seminar, "What You Hear is What You See?" where faculty, scholars, and researchers from more than 20 institutions examined the integration of data sonification and visualization.

In addition to these discussions, a hands-on activity invites students, faculty, and practitioners to explore how soundscape analysis can support sustainability in the textile industry. While Life Cycle Assessment (LCA) offers critical environmental insights, it is a slow and retrospective process. By contrast, machine learning-driven sound analysis can provide real-time diagnostics, identifying inefficiencies in manufacturing by detecting patterns in machinery and production noise. It is well known that the textile industry consumes large amounts of water and energy and, consequently, generates wastewater and energy waste that requires comprehensive intervention. Circularity, therefore, becomes a necessary avenue to achieve the minimization of the planet's finite resources waste while guaranteeing the social and financial sustainability of the sector. For this purpose, the state-of-the-art tool applied is the Life Cycle Assessment (LCA) and the related water, carbon or energy footprints, which provide a diagnosis of the situation and allow the proposal of environmental (and, consequently, socioeconomic) improvements. However, LCA is a complex and asynchronous process that requires data collection over a long period of time. While many industries carry out these assessments on a regular basis, they are still far from providing real-time information which would dramatically increase the stakeholders’ agency and the possibility to optimize water and energy consumption - and the consequent footprint - of a production facility. Recent research has shown that the analysis of the soundscape in an indoor environment (i.e. the sounds produced by the different agents such as humans or machines) are a valuable source of information: for instance, the sound of an engine can inform listeners on its correct functioning. If automatized by leveraging ML techniques, soundscape analysis can provide users with insights on the ‘auditory footprint’ of sound events which, conversely, provide information on the actions that are behind the sound - for instance, the different steps of a production chain. During this activity at the DRW we invite participants to brainstorm - starting from a collection of sound recordings from a shoe manufacturing plan - on the ‘autographic’ value of audio data, potential technological solutions that include Machine Listening techniques, other application scenarios, and the relationship between data, sound, and sustainability.

Host or Publisher Center for Design at CAMD