To increase energy-efficiency and reduce CO2e emissions in maritime shipping, Decision-Support Systems (DSS) can be leveraged. Specifically, in regard to reducing the greatest contributor to consumption, propulsion (IMO, 2021), by assisting seafarers in route planning, and timely and efficient re-planning, as well as general monitoring of ship’s energy dynamics. However, the successful integration and acceptance of these systems into the seafarer’s workflow pose significant challenges, such as goal conflicts, e.g. with safety or with the financial interests of different stakeholders, which require a deep understanding of interactions onboard and onshore.This paper reflects on our implementation of a transdisciplinary design research approach for developing novel, human-centered AI-based tools for energy-efficient ship operations. Of our concurrent studies, we describe selected forms of inquiry that together resulted in a holistic understanding of the application domain, target audience, and typical tasks as well as an interactive prototype of a decision support system for energy-efficient ship navigation.The research activities reported are based on human factors research concerning energy-efficient ship operations and focus on research through design in the sense of Jonas (2015) in the field of DSS for CO2e emission mitigation in navigation and ship operation, and the formative evaluation of a DSS prototype in a ship simulator environment (N = 22). By viewing these research activities through the lens of design research, more specifically the theoretical foundation of MAPS (Jonas et al., 2010), we systematically describe and discuss their individual contributions. MAPS specifically operationalized design research as “Matching Analysis, Projection and Synthesis”, enabling integrative, systematic research processes across boundaries of disciplinary bodies of knowledge, domains and actors.As a primary contribution, we reflect on our lessons learned to identify generalizable challenges for similar future projects of the maritime ergonomics community. These include (1) context-sensitive integration of navigational and operational data; (2) calibration of users’ expectations of the system’s capabilities; and related to this (3) increasing transparency of how the DSS retrieves and processes data, and of how confident it is in its suggestions. By considering key human factors, such as workload, autonomy and biases (e.g., automation bias) on the basis of our system, we demonstrate how these challenges can be addressed. As a secondary contribution, we also share our resulting designs as examples of how AI-based decision support for optimizing energy efficiency can be visually and functionally integrated into onboard ship operation and navigation.