10 September 2026
As artificial intelligence moves beyond controlled settings and into dynamic, uncertain and socially complex environments, new questions arise about how these systems understand their surroundings, act autonomously, and interact with people.
Rather than concentrating intelligence in a single complex system, the four-year European research project EMERGE took a different approach. It considered the simplest possible artificial agents, such as small robots in a swarm, components of a robotic body, or connected devices. Individually, each unit may possess limited intelligence and information. Together, however, from their interactions, sharing information and coordinating their behaviour, they can establish a representation of shared existence, environment and goals, which the project calls collaborative awareness.
During the project, an interdisciplinary group of researchers in artificial intelligence, robotics, mathematics, philosophy and cognitive science from the University of Pisa (IT), Ludwig Maximilian University of Munich (DE), Delft University of Technology (NL), the University of Bristol (UK) and Da Vinci Labs (FR) worked together to develop a philosophical, mathematical and technological framework for this collaborative awareness in artificial systems.
The EMERGE project begins by establishing a fundamental distinction between awareness and consciousness. Consciousness is generally associated with subjective or phenomenal experience, with having feelings, and experiencing the world from a particular perspective.
“In EMERGE, we understand awareness more operationally, as the capacity of an agent to process and integrate information in a way that is relevant to its actions,” explains Ophelia Deroy, philosopher and cognitive scientist at LMU Munich. And our experiments showed that people can actually understand an artificial system as aware without assuming that it has subjective experience.”
To further establish that distinction, the researchers examined awareness through different dimensions that can be applied across individuals and collective systems, including spatial, temporal, self, agentive and metacognitive awareness. Rather than treating awareness as an all-or-nothing property, the framework connects each dimension to specific capabilities and tasks. This makes it possible to test whether increasing a system’s awareness improves its performance.
“Our work also examined the ethical side of human interaction with collaboratively aware systems,” says Bahador Bahrami, director of the Crowd Cognition group at LMU Munich, “with important implications for future environments in which groups of humans and artificial agents will have to interact, negotiate and cooperate.”
Their studies explored, for example, how people behave toward automated systems such as self-driving cars and found that users may be more willing to take advantage of artificial agents than human counterparts, a phenomenon described by the researchers as algorithmic exploitation. Other findings also inform the design and governance of artificial systems that interact directly with people.
Furthermore, the consortium partners developed mathematical and computational tools to describe how awareness can emerge across a physically distributed systems by exchanging simple local information with neighboring agents and integrating it to coordinate their actions toward solving a task.
“The beauty of the approach is that the system operates in a distributed way,” says Sabine Hauert, professor at the University of Bristol. “This way systems that can scale to large numbers of agents and remain operational when individual robots fail or conditions change. It also allows humans to interact with the systems as a coordinated collective rather than controlling each robot separately.”
“We want robots that can understand how their actions affect the world, anticipate what people and other robots are doing, and adapt accordingly. That is a key step toward robotic systems that can operate more autonomously and work naturally alongside humans,” explains Cosimo Della Santina, associate professor at TU Delft.
Moreover, the consortium’s mathematical framework also showed potential for more parsimonious machine learning, with systems capable of learning from substantially smaller quantities of data than conventional architectures.
“Many AI systems today depend on large, centralized computing infrastructures. Our results open the way to a more decentralized and environmentally sustainable models, in which data can be processed locally using small and energy-efficient devices, such as neuromorphic hardware, reducing energy consumption, network traffic and latency while improving privacy.” says Claudio Gallicchio, associate professor at the University of Pisa.
EMERGE was designed to establish a pathway from foundational research to technological innovation and, as such, the project’s results are already supporting several spinout initiatives: Embodied AI, which develops efficient AI models for future generations of humanoid robots; ContinualIST which develops more efficient AI models that can run on smaller, lower-power hardware; Collective Robotics, focused on swarms of robots that can coordinate and operate collectively; and Phoebe, dedicated to systems in which humans and artificial agents work together to solve problems.
The concepts of collective collaboration and awareness developed in EMERGE will also be central to new research projects. One of them (EIC Pathfinder SWIFT-BUILD) will apply the concept of collaboration among multiple robots and humans to automated construction, while two others (HE MINGLE and ADA-COLLAB) investigate how general-purpose, agentic AI systems can form societies, collaborate with one another, and interact with humans.
Additionally, EMERGE developed nOdes, an artistic installation that allows the public to experience collaborative awareness directly, through an interactive swarm whose agents exchange information and change their behaviour in response to one another and to human participants.
“EMERGE started from a fundamental question about awareness, but it has ended with very concrete results: new robotic systems, more sustainable AI architectures, new startup initiatives and follow-up projects that are already carrying forward the idea of building artificial systems that are more efficient, sustainable, collaborative, and imbued with human values,” concludes Davide Bacciu, professor at the University of Pisa and coordinator of the project.
The EMERGE project was funded by the European Union under Grant Agreement 101070918. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Innovation Council and SMEs Executive Agency (EISMEA). Neither the European Union nor the granting authority can be held responsible for them. UK participants in Horizon Europe Project 101070918 are supported by UKRI grant number 10038942.
