Overview

Multi-robot systems (MRS) are a sub-class of multi-agent systems (MAS) which have great potential as regards parallelism, efficiency and robustness in tasks which are either distributed in space or too complex to be accomplished by a single mobile robot. These benefits hinge on suitable internal team organization to achieve coordination and cooperation, which in turn requires efficient sharing of information. MRS may be very useful on assisting humans in many distributed activities, especially in hazardous scenarios, by extending human perception and actuation with distributed sensors and actuators. The CHOPIN R&D project aims at exploiting this human-robot symbiosis in the development of human rescuers’ support systems for small-scale search and rescue missions in urban catastrophic incidents, an application domain with an unquestionable beneficial impact on society. A proof of concept will be developed for innovative techniques about cooperation between teams of human agents and teams of mobile robotic agents and collaborative context awareness.

The scope of security operations has been changing rapidly for the past few years. Nowadays, the threats for citizens of developed countries arising from classical military conflicts are decreasing but, on the other hand, there has been a strong increase in the need to respond effectively to catastrophic, unexpected incidents, including natural and civil disasters (e.g. collapses, fires, floods and earthquakes), industrial accidents or technological disasters (e.g. accidents in nuclear reactors, refineries, etc.), and terrorism acts and crime (e.g. bomb attacks). Current security organizations are facing shortage in specialized dedicated equipment, which leads to extreme exposure to risk of human lives within search and rescue teams and less than desirable effective victim assistance within civil population.

From the standpoint of economic and social development, this project will contribute to increase the safety of human rescuers and potential victims, and to attain faster, more precise, and cost effective response to catastrophic incidents, through the use of teams of mobile robotic agents in cooperation with teams of human rescuers, a.k.a. first responders in the context of security operations. Robotic technology can support first responders in most dangerous or repetitive actions during crisis operations, by taking advantage of robots’ expendability. It enables rescue teams to concentrate resources on core activities without or with limited stress arising from life endangerment. Dealing with terrorist bomb threats in public areas, or assessing infrastructural conditions in old buildings (e.g. in historic areas of cities) or in buildings at risk of ruin or after collapsing (e.g. because of a fire or seismologic phenomena), are examples of scenarios that will be used to validate the project’s scientific contributions. The human-robot system to be studied within this project will comprise basically hand-held devices (HHD) to be worn by first responders and provide them with effective and augmented situation awareness, and mobile robots (MR) which assist first responders in performing security missions.

Cooperation in MRS has been deeply studied by roboticists for the past two decades. However, there is still a gap on studying the cooperative interaction between teams of cooperative mobile robots and teams of humans, which this project aims to exploit. Although the study of cooperation in the interface of those teams can resemble some problems of MAS in the scope of Artificial Intelligence, several challenging scientific problems remain unsolved when that cooperation must take place in real implementations with human and robotic agents. The most important one is the efficient sharing of information through information utility assessment, as massive inter-agent communication is impractical, either because of communication channel’s limited bandwidth or overwhelming information processing on each agent. Another relevant scientific problem is collaborative context awareness and context sharing between team of humans and team of robots. Human and robotic agents must be able to create a common interface to attain a correct and common interpretation of shared information given the current shared context. Building global awareness from local awareness, i.e. imprecise and incomplete information, and developing context recognition techniques for robotic devices are other subsidiary interesting issues.

In summary, this project will focus on the following scientific goals:

• SO1.Architectures and models for cooperation in teams of humans and in teams of mobile robots;
• SO2.Collaborative context awareness and context sharing between team of humans and team of robots;
• SO3.Scalable and cooperative sharing of information based on measures of information utility assessment.