Towards a Policy Development Methodology for Human-Centred IoT Collectives

Area of Science:

• Computer Science
• Software Engineering
• Artificial Intelligence Ethics

Background:

• Smart Internet-connected devices require ethical considerations for human-centered operation.
• Current software industry practices lack robust methodologies for embedding socio-ethical policies into smart devices.
• Ensuring socially acceptable behavior in smart devices is a growing challenge.

Purpose of the Study:

• To propose a novel policy development methodology for enhancing the human-centered behavior of smart devices.
• To provide a structured approach for defining smart device roles and aligning them with socio-ethical and administrative policies.
• To demonstrate the applicability and effectiveness of the proposed methodology across diverse interaction settings.

Main Methods:

• A four-phase methodology was developed: concept development, policy definition and mapping, policy processing implementation, and device deployment.
• The methodology was applied and tested in three distinct real-world scenarios: a supermarket, a children's hospital, and early learning centers.
• The methodology was implemented on various smart internet-connected devices, including robots, smart cameras, and smart speakers.

Main Results:

• Smart devices demonstrated significantly more human-centric behavior after the methodology's application.
• The implemented devices successfully adhered to defined socio-ethical policies during their core functions.
• The methodology proved effective in diverse environments requiring human-device interaction.

Conclusions:

• The proposed policy development methodology effectively promotes human-centered and ethically compliant behavior in smart devices.
• The framework offers a scalable and adaptable solution for integrating ethical considerations into the design and deployment of smart technologies.
• Further research can explore expanding this methodology to a wider range of smart devices and complex social contexts.