AMaLIA-2019-JYU (AMaLIA)
Main funder
Funder's project number: 3736/31/2019
Funds granted by main funder (€)
- 108 000,00
Funding program
Project timetable
Project start date: 01/08/2019
Project end date: 30/09/2020
Summary
Advanced Machine Learning for Industrial Applications (AMaLIA-2019) is composed of 5 Work
Packages within the NSF-BF CVDI 2019-2020 Finland Site Projects, approved by NSF CVDI
Steering Group. The main target of AMaLIA is to develop advanced Artificial Intelligence (AI) and
Machine Learning techniques and apply them in relevant, timely, and high impact industrial
applications of interest to the Industry Advisory Board and the industries they represent. The
project is organized in 5 WPs. WP1 deals with real-time data collection, anomaly detection,
localization and classification at the edge by taking into acount energy efficiency and
privacy-preservation. The goal of this project is to enable visual analytics on data streams at the
edge. WP2 develops a real-time health monitoring system, that is energy efficient and
privacy-preserving. The novel goal of this project is to design a domestic body sensor network
(BSN) prototype that is able to simultaneously aggregate multiple measurements in a compressed
manner, which is also privacy preserving at low cost. WP3 deals with the develoment of advanced
machine learning algorithms for anomaly detection. The aim of the WP is to develop novel
Autoencoders and Variational Autoencoders to detect specific types of anomalies, such as cracks
and potholes on road surfaces in real-time. WP4 aims to push the current state of the art in
machine learning by developing Self-taught and semi-supervised learning and apply them in
various applications of interest to the sponsoring company in CVDI. WP8 produces a
Proof-of-concept (PoC) for Statistical modeling and visualization of human cognition and strategic
business data. It applies artificial intelligence and machine learning to two diverse data sets – one
dealing with the room design, people interaction and cognitive stiumuls and other dealing with
strategic business data, applied in business development context.
Packages within the NSF-BF CVDI 2019-2020 Finland Site Projects, approved by NSF CVDI
Steering Group. The main target of AMaLIA is to develop advanced Artificial Intelligence (AI) and
Machine Learning techniques and apply them in relevant, timely, and high impact industrial
applications of interest to the Industry Advisory Board and the industries they represent. The
project is organized in 5 WPs. WP1 deals with real-time data collection, anomaly detection,
localization and classification at the edge by taking into acount energy efficiency and
privacy-preservation. The goal of this project is to enable visual analytics on data streams at the
edge. WP2 develops a real-time health monitoring system, that is energy efficient and
privacy-preserving. The novel goal of this project is to design a domestic body sensor network
(BSN) prototype that is able to simultaneously aggregate multiple measurements in a compressed
manner, which is also privacy preserving at low cost. WP3 deals with the develoment of advanced
machine learning algorithms for anomaly detection. The aim of the WP is to develop novel
Autoencoders and Variational Autoencoders to detect specific types of anomalies, such as cracks
and potholes on road surfaces in real-time. WP4 aims to push the current state of the art in
machine learning by developing Self-taught and semi-supervised learning and apply them in
various applications of interest to the sponsoring company in CVDI. WP8 produces a
Proof-of-concept (PoC) for Statistical modeling and visualization of human cognition and strategic
business data. It applies artificial intelligence and machine learning to two diverse data sets – one
dealing with the room design, people interaction and cognitive stiumuls and other dealing with
strategic business data, applied in business development context.
Principal Investigator
Primary responsible unit
Profiling area: Cyber security (University of Jyväskylä JYU)
