<p>Existing studies on indoor position recognition employ diverse evaluation methods, which complicates direct accuracy comparisons across techniques. To address this issue, this study proposes a novel framework for evaluating the accuracy of indoor position recognition methods. The proposed framework evaluates accuracy by using the position recognition results of a grid-pattern-tracking autonomous mobile robot (GPT-AMR) as a benchmark. To validate the proposed evaluation method, a comparative analysis was conducted on four position recognition algorithms: (1) a computer vision-based algorithm, (2) a Bluetooth Low Energy (BLE)-based trilateration algorithm, (3) a BLE-based adaptive trilateration algorithm, and (4) a least squares method (LSM)-based algorithm. Experimental results demonstrated that the proposed evaluation method, which employs GPT-AMR, offers improved speed, accuracy, and practical applicability compared to conventional approaches. Furthermore, this method enables objective comparisons and evaluations of a wide range of indoor position recognition technologies, including both computer vision- and BLE-based algorithms, using a standardized criterion. Future research will focus on systematically validating the generalizability of the proposed method across different indoor environments and operational conditions. This study aims to advance indoor position recognition technology for autonomous mobile robots (AMRs) and improve their applicability in various service robotics domains.</p>