Recently, the embodied AI enterprise Shentingji officially introduced SWAN, a specialized architecture tailored for outdoor companion robots, designed to function in real-time at the edge. SWAN comprises three key modules: a model inference backbone, a hybrid memory identity database, and a secure data flywheel system. These components collectively empower the architecture with real-time decision-making abilities, long-term memory retention, and secure iterative functionality. According to reports, SWAN will initially be integrated into Shentingji's debut consumer-grade two-wheeled-legged outdoor companion robot, Rovar, which is slated for release by the year's end. Furthermore, Shentingji has successfully secured its Pre-A round of financing, marking a significant milestone in its growth trajectory.
