9a.006.JYU - Beyond PoC: Ethically aligned design, measurement and visualization of human interaction and business strategy execution (WP8 & WP9)

Project - Summary

Physical workspace can have a notable impact on the work performance of the employees in any given organization. By understanding human cognition, we can also seek to understand how the workspace can positively or negatively affect work performance, specifically from the point of view of workplace interaction performance.

Business consultants work in a very data intensive field. In order to gain an understanding of their clients’ business and help them solve the problems and challenges they are having consultants first need to gather data. After the data gathering, strategic choices to overcome the problems clients are facing are made. One approach for strategy creation is to utilize a Strategic Analysis model to identify the most crucial goals for the company to address to improve their business performance – and to define means for achieving them.

In this Beyond PoC phase our focus was on utilizing actual data gathered from office environment, applying machine learning to human interaction behavior analysis, further developing the experimental setup from the CVDI year 8 project, developing means for the visualization and reporting of the data, and overall improving the workplace interaction performance through findings derived from the experiments.

Other part of this project focused on building a Knowledge-Based System (KBS) utilizing Natural Language Understanding/Processing (NLU/NLP) for guiding the user through the Strategic Analysis process of identifying the most crucial goals and defining means to achieve them for improved business performance.

Project - Team

Team Member Role Email Phone Number Academic Site/IAB
Pekka Abrahamsson PI 358 40 541 5929 University of Jyväskylä
Kai-Kristian Kemell Researcher N/A University of Jyväskylä
Ville Vakkuri Researcher N/A N/A University of Jyväskylä
Juhani Risku Researcher N/A N/A University of Jyväskylä
Joonas Himmanen Student 358 44 062 9059 University of Jyväskylä
Antti Kariluoto Student 358 44 285 2431 University of Jyväskylä
Joni Kultanen Student 358 50 375 8568 University of Jyväskylä
Mikael Ovaska Student N/A University of Jyväskylä
Teemu Autto Student N/A University of Jyväskylä
Joonas Uusnakki Student 358 44 296 6342 University of Jyväskylä
Mikko Virtaneva Project Mentor 358 44 703 9200 Workspace Oy
Jukka Lassila Project Mentor 358 50 1810 Multimentor Oy
Funded by: Workspace Oy & Multimentor Oy

Project - Novelty of Approach

Human interaction and performance, and business strategy execution data both present new domains for Machine Learning applications. The problem is approached from ethics viewpoint by using ECCOLA – a Method for Implementing Ethically Aligned AI Systems [1] throughout the design and development of the AI powered PoC solutions. While research conducted on performance and productivity in a remote work setting has increased due to the COVID-19 pandemic [e.g., 2, 3, 4, 5], our study distinguishes by the longitudinal data we have been able – and will continue – to gather from the beginning of the pandemic through changes in work customs – both in remote and office settings in different phases of the pandemic. Examining the factors identified in existing literature to be affecting individual performance and productivity – such as work-related stress [6, 7, 8], teamwork [9, 10, 11], and self-efficacy [12, 13, 14] – through longitudinal data provides unique view into how the pandemic may change the perception of work as a whole and how the significance of the previously identified factors to performance and productivity might differ from the past after the prolonged period of remote work.

Project - Deliverables

1 Empirical test environment -report
2 Human interaction models and performance -report
3 Business strategy execution models and performance -report
4 Empirical evidence of ethically aligned design -report

Project - Benefits to IAB

We have submitted three papers for this project to the 1st Workshop on Requirement Engineering for Software startups and Emerging Technologies (RESET 2021) – part of the IEEE International Requirements Engineering Conference 2021 (RE2021) – titled:

  • “Visualizing Interactions in Workplace Office: Key Lessons Learned from an Experiment”
  • “Web Bluetooth: Experiences in Implementing Emerging technology to Legacy System”
  • “Deep Neural Network Voice Activity Detector for Downsampled Audio Data: An Experience Report”

Paper with title: “COVID-19 Remote Work: Body Stress, Self-Efficacy, Teamwork, and Perceived Productivity of Knowledge Workers” [15] from the preceding CVDI year 8 project was accepted and will be presented at the 12th Scandinavian Conference on Information Systems (SCIS 2021).

In addition to the articles, a master’s thesis with title “Current state of practice on AI ethics implementation in the industry” [16] was enabled by this CVDI year 9 project.

Project - Documents

projects/year9/9a.006.jyu.txt · Last modified: 2021/08/24 09:14 by sally.johnson