Case Study : Emotional AI in Maritime Mobility: Understanding Communications to Support Decisions on Board

On board a vessel, the quality of communication matters as much as the information being shared. An instruction may be understood yet still cause hesitation. A message may convey concern without stating it explicitly. In a tense situation, these nuances can help reveal what is happening within a crew.

Emoticonnect develops an emotional intelligence layer that integrates into existing tools and embedded systems. In the maritime sector, this approach creates opportunities to better understand human interactions, adapt interfaces and support decision-making.

Adding an Emotional Dimension to Available Information

Navigation and monitoring systems provide essential data about a vessel and its surroundings. Emotional AI adds complementary information: the emotional signals expressed in communications.

Using sources agreed upon with the client, such as voice communications or text messages, Emoticonnect analyses emotional cues and how they evolve within an interaction.

The aim is to make certain signals more visible to those making decisions or supporting a situation. Increasing tension during a conversation, recurring concern or expressed confusion can draw attention to an exchange and prompt a human check.

These results provide contextual indicators. On their own, they cannot establish a person’s psychological condition, fitness for duty or the presence of danger.

Identifying Tension in Onboard Communications

A first use case involves exchanges between crew members, or between a vessel and shore teams.

When several people are involved in a situation, information may circulate rapidly, under time pressure and in difficult communication conditions. Emotional analysis can help identify moments when exchanges become more tense or require clarification.

For example, if a conversation reveals signs of concern alongside repeated requests for confirmation, the system can suggest rephrasing an instruction or checking that it has been understood. The person responsible retains control over assessing the situation and deciding what to do.

This approach aims to facilitate coordination and support collective awareness.

Adapting Interactions Between Crews and Embedded Systems

A second use case involves onboard interfaces, including voice assistants, monitoring tools and decision-support systems.

By integrating the Emoticonnect layer, an interface can take account of emotional cues expressed by the user and adapt its responses. Depending on the chosen scenario, it could offer a clearer explanation, present information more concisely or request confirmation before proceeding.

Consider a crew member questioning an onboard assistant and expressing concern through their questions. The assistant could adjust its wording and suggest a relevant clarification, while following established procedures.

The aim is to design interactions that better account for how people receive and use information, particularly in demanding situations.

Supporting Decisions and Contributing to Seafarer Safety

Emotional analysis becomes useful when it forms part of a clearly defined operational process.

An isolated indicator offers limited value. A signal connected to the context of an exchange, however, can help focus attention: should an instruction be clarified, information checked or another person brought into the discussion?

Emoticonnect can deliver these indicators and related recommendations through the tools teams already use. How the information is presented is defined according to each person’s responsibilities and the needs of the use case.

Its contribution to seafarer safety must then be evaluated in the field. This depends on the reliability of the analysis in maritime conditions, the relevance of recommendations and their practical usefulness to crews.

Sovereign Integration Within the Client’s Environment

Emoticonnect integrates via API into existing platforms and equipment. This enables emotional analysis capabilities to be added to a communication system, monitoring interface or AI assistant.

For environments requiring sovereign processing, deployment can take place on the client’s infrastructure or directly on board. In this configuration, no data leaves the client’s ecosystem.

The scope of analysis, access to results and retention rules are defined in advance. The project must also establish how the people concerned are informed and which uses of the analysis are authorised.

Developing and Evaluating Applications with Maritime Professionals

Maritime environments present specific challenges: background noise, radio communications, specialist terminology, different languages and accents, and limited connectivity. These conditions must be considered from the outset.

A pilot allows partners to select a concrete scenario, identify available sources and agree on success criteria. Evaluation can cover the quality of analysis in real operating conditions, the relevance of recommendations and the system’s ability to avoid unnecessary alerts.

User feedback is essential to determine whether the information genuinely helps people in their work. Deployment at scale follows from these findings.

This approach can support both civilian and military applications, with requirements tailored to each environment.

Brest: A Relevant Base for Developing These Applications

At Sea Tech Week in Brest, our CEO and founder, Marie Argence, presented these perspectives during a workshop focused on the region’s attractiveness.

A scientific ecosystem that includes UBO, École navale, Lab-STICC and Ifremer, alongside local expertise in maritime innovation and defence, makes Brest a relevant location for exploring these applications.

These discussions strengthened our commitment to working with partners who share this ambition: developing embedded emotional AI that supports people, integrates into their tools and is evaluated under their operating conditions.

Are you developing an embedded system, a monitoring interface or a maritime communication solution? Contact Emoticonnect to define an initial use case and its evaluation framework together.