Connected objects (CO), whether medical devices or not, are used in clinical research for data collection, a specific activity (communication, diagnosis, effector, etc.), or several functions combined. Their validation should be based on three approaches: technical and clinical reliability, data protection and cybersecurity. Consequently, the round table recommends that the typology of COs, their uses and limitations, be known and shared by all, particularly for implementing precise specifications. COs are used in clinical research during observational studies (assessment of the device itself or data collection), randomized studies, where only one group has a CO (assessment of its impact on patient follow-up or management), or randomized studies where both groups have a CO, which is then used as a tool to help with assessment. The benefits of using COs in clinical research includes: improved collection and quality of data, compliance of patients and pharmacovigilance, easier implementation of e-cohorts and a better representative balance of patients. The societal limits and risks identified relate to the sometimes intrusive nature of certain collected parameters and the possible misuse of data. The round table recommends the following on this last point: anticipation, by securing transmission methods, the qualification of data hosts, and assessment of the object's vulnerability. For this, a risk analysis appears necessary for each project. It is also necessary to accurately document the data flow, in order to inform both patients and healthcare professionals and to ensure adequate security. Anticipating regulatory changes and involving users starting from the study design stage are also recommended.