This study aimed to map the landscape of robotic endoscopic platforms developed to support Endoscopic Submucosal Dissection (ESD) and to describe how these systems have been evaluated across preclinical and early clinical settings. It also sought to assess their technological maturity using the Technology Readiness Level (TRL) and IDEAL frameworks. In addition, the review aimed to synthesise reported outcomes comparing robot-assisted (R-ESD) with conventional ESD (C-ESD) in preclinical studies and to evaluate emerging evidence from early clinical investigations. 

A scoping review was conducted and reported according to PRISMA-ScR. A comprehensive search of Ovid MEDLINE, Embase, Cochrane Library, Web of Science, Scopus, and IEEE Xplore was performed in February 2025 and updated in October 2025. Studies evaluating robotic endoscopic platforms for ESD in preclinical porcine models, where the experimental setup adequately replicated the procedural complexity of ESD, or in clinical settings, were included. Two reviewers independently screened records, and disagreements were resolved through consensus. Data were charted using a standardised extraction form. Extracted variables included study characteristics, robotic platform features, experimental design, and procedural outcomes. Findings were synthesised descriptively using both narrative synthesis and tabular mapping. 

A total of 27 studies were ultimately included. The studies were published between 2010 and 2025, with most appearing from 2019 onwards (22; 82%), and originated mainly from China (9; 33%) and South Korea (8; 30%). Most studies were preclinical (21; 78%), including 8 ex vivo (30%), 11 in vivo (41%), and 2 combining both models (7%). Six studies were clinical (22%), with only one being a multicentre randomised controlled trial (RCT). Over half were comparative (15; 56%), predominantly evaluating R-ESD vs C-ESD (14; 52%). The most frequently studied anatomical site was the stomach (18; 67%). The most frequently evaluated platforms were the REXTER/ROSE/ROBOPERA series (6; 22%), followed by the MASTER/EndoMaster system (5; 19%) and the FASTER/EndoFaster platform (4; 15%). Several other systems (n = 8) were each evaluated in one to two studies. Most platforms were classified within TRL 4–7, having been validated in ex vivo or in vivo models. A smaller number—namely ROBOPERA, EndoMaster, FLEX, and EndoFaster—reached TRL 8–9, reflecting progression to clinical evaluation. Thirteen preclinical studies were included in the comparative synthesis. Among studies reporting these outcomes, 11/12 (92%) showed shorter submucosal dissection times with R-ESD, and all reporting submucosal dissection speed (8/8; 100%) favoured R-ESD over C-ESD. En bloc resection rates were high overall, with nine studies (70%) achieving 100% in both groups, and lower rates occurring mainly in novice-performed C-ESD. Perforation was uncommon in expert-led studies, with R-ESD consistently reporting 0%, while higher perforation rates occurred predominantly during novice-performed C-ESD. Six clinical studies were identified across IDEAL Stages 1–3, and early feasibility work demonstrated successful en bloc and complete (R0) resections with acceptable safety profiles.  

Robot-assisted ESD is an emerging field with a growing number of experimental platforms and increasing early-phase clinical activity. Preclinical studies consistently demonstrate technical feasibility and potential performance advantages of R-ESD over C-ESD, particularly in dissection speed and safety under controlled conditions. Early clinical investigations indicate the feasibility and safety of R-ESD, although clinical use remains limited to early-phase studies. Overall, robotic platforms for ESD show promise, but further development, standardisation, and robust clinical evaluation are needed before wider adoption can be realised.