Best Practices for Developing Socially-Aware Service Robots with Predictive Techniques to Enhance Social Comfort

Over the last few years, there has been a growing interest in the development of service robots that can operate in everyday environments, both in commercial and research settings. These robots are designed to work autonomously in busy settings. They aim to move considering efficiency and safety criteria, but also social acceptance. Incorporating social awareness criteria into the conventional path planning modules, while maintaining speedy algorithms that can adjust to human behavior without causing discomfort, presents a significant challenge. To overcome this challenge, predictive methods have become increasingly relevant. It is also important to establish guidelines for developing methods that promote social acceptance. This includes considering factors such as cultural differences, ethical considerations, and user feedback to ensure that the robots are accepted and trusted by society. This paper compares different state-of-the-art predictive methods in the field of socially aware robot navigation. It also proposes guidelines to evaluate and train these methods, taking into account social acceptance as a key requisite to successfully deploy service robots in daily life settings.