In recent years, commercial and research interest in service robots working in everyday environments has grown. These devices are expected to move autonomously in crowded environments, maximizing not only movement efficiency and safety parameters, but also social acceptability. Extending traditional path planning modules with socially aware criteria, while maintaining fast algorithms capable of reacting to human behavior without causing discomfort, can be a complex challenge. Solving this challenge has involved the development of proactive systems that take into account cooperation (and not only interaction) with the people around them, the determined incorporation of approaches based on Deep Learning, or the recent fusion with skills coming from the field of human–robot interaction (speech, touch). This review analyzes approaches to socially aware navigation and classifies them according to the strategies followed by the robot to manage interaction (or cooperation) with humans.

Social robots, designed to interact and assist people in social daily life scenarios, require adequate path planning algorithms to navigate autonomously through these environments. These algorithms have not only to find feasible paths but also to consider other requirements, such as optimizing energy consumption or making the robot behave in a socially accepted way. Path planning can be tuned according to a set of factors, being the most common path length, safety, and smoothness. This last factor may have a strong relation with energy consumption and social acceptability of produced motion, but this possible relation has never been deeply studied. The current paper focuses on performing a double analysis through two experiments. One of them analyzes energy consumption in a real robot for trajectories that use different smoothness factors. The other analyzes social acceptance for different smoothness factors by presenting different simulated situations to different people and collecting their impressions. The results of these experiments show that, in general terms, smoother paths decrease energy consumption and increase acceptability, as far as other key factors, such as distance to people, are fulfilled.

Service robotics involves the design of robots that work in a dynamic and very open environment, usually shared with people. In this scenario, it is very difficult for decision-making processes to be completely closed at design time, and it is necessary to define a certain variability that will be closed at runtime. MAPE-K (Monitor–Analyze–Plan–Execute over a shared Knowledge) loops are a very popular scheme to address this real-time self-adaptation. As stated in their own definition, they include monitoring, analysis, planning, and execution modules, which interact through a knowledge model. As the problems to be solved by the robot can be very complex, it may be necessary for several MAPE loops to coexist simultaneously in the robotic software architecture endowed in the robot. The loops will then need to be coordinated, for which they can use the knowledge model, a representation that will include information about the environment and the robot, but also about the actions being executed. This paper describes the use of a graph-based representation, the Deep State Representation (DSR), as the knowledge component of the MAPE-K scheme applied in robotics. The DSR manages perceptions and actions, and allows for inter- and intra-coordination of MAPE-K loops. The graph is updated at runtime, representing symbolic and geometric information. The scheme has been successfully applied in a retail intralogistics scenario, where a pallet truck robot has to manage roll containers for satisfying requests from human pickers working in the warehouse.

In the past few years, there has been an increase in commercial and research focus on service robots operating in daily surroundings. These machines are anticipated to function independently in busy settings, enhancing movement efficiency and safety parameters, as well as social acceptance. Expanding conventional path planning modules to include socially aware criteria, while sustaining speedy algorithms that can adapt to human behavior without causing distress, presents a significant challenge. To address this challenge, learning methods have gained significant relevance. Among the various techniques, deep reinforcement learning, end-to-end, and inverse reinforcement learning have been the most promising. However, it is difficult to determine which techniques are superior, and sometimes, developers may obtain poor results due to inadequate data or experimental procedures during the learning stage. Therefore, it is essential to evaluate and discuss the best practices and options for an effective training stage that can improve results. As we are specifically referring to social robots, the evaluation of results should take into consideration social comfort as a key factor.