The next decade is going to witness a profound transformation in the mobile communication infrastructure. The aim of interconnecting everything everywhere unleashes new visions such as autonomous vehicles, factory of the future, and augmented reality. To meet the requirement associated with these new use cases, one key technology is the millimeter-wave (mmWave) and sub-Terahertz (sub-THz) spectrums. However, operating at such high frequencies remains challenging, especially in mobile scenarios. Indeed, mmWave operation includes a beam training process to align the directional beams radiated by the antennas at the base station and the user equipment. This procedure introduces large overhead signaling as well as latency, especially during initial access, which could jeopardize mmWave effectiveness. So far, beam training relies on the sequential scanning of the beam space, which represents a bottleneck for efficient mobile communications. To address the latency and overhead issues in this beam management, BeSensiCom conducts a radically different approach by envisioning a sensing unit that feedbacks the angular spectrum to the base station in real time. To achieve such a technological breakthrough, BeSensiCom leverages the intrinsic frequency beam scanning behavior of multibeam leaky-wave antennas to perform angle-of-arrival estimation using dedicated algorithms. BeSensiCom goes beyond the state-of-the-art by conducting a transdisciplinary approach combining applied electromagnetics and signal processing. With BeSensiCom’s vision, the cognitive approach of frequency spectrum sensing is extended to the angular spectrum to enable managing the beam space not as a separate resource, but as an extra dimension in the resource grid. This brings the unique opportunity to represent channel resources, not as a 2D grid, but as a time-frequency-beam 3D grid.