Robotic technologies in prosthodontics and implantology: a scoping review of current applications and future directions

Abstract

Purpose

This scoping review synthesizes current evidence on the applications, effectiveness, workflow, and limitations of robotic technologies in prosthodontics and implant dentistry, focusing on their potential to enhance precision, reproducibility, and clinical outcomes.

Materials and methods

A systematic search was conducted across PubMed, SCOPUS, EMBASE, and Web of Science from January 2015 to May 2025, following PRISMA-ScR guidelines. Studies were screened and selected based on predefined inclusion criteria, with data extracted into categories of prosthetic dentistry, implant dentistry, and robotic development stages.

Results

Twenty-one studies were included (7 prosthodontics, 14 implantology). In prosthodontics, robotic systems demonstrated high accuracy in veneer and crown preparations (errors < 0.4 mm) and automated denture arch generation (positioning accuracy 0.09 mm). Implant dentistry studies revealed robotic systems (Yomi, Remebot) outperformed static/dynamic navigation in precision, with mean angular deviations < 2.81° and coronal/apical deviations < 1 mm. Semi-active and active robots dominated implant applications, while prosthodontic robots were primarily active but less clinically advanced.

Conclusions

Robotic technologies hold transformative potential for implant and prosthetic dentistry, with higher precision and workflow efficiency. However, robust clinical trials and cost-effective solutions are needed for adoption of these systems in regular clinics.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-026-08506-0.

Introduction

The integration of precision technologies in modern dentistry has significantly improved clinical outcomes, particularly in implant and prosthetic procedures. Dental implant surgery, in particular, demands accuracy in the placement of implants with respect to angulation, depth, and positioning within the alveolar bone [1]. Any deviation from the planned treatment can result in complications such as nerve injury, excessive bleeding, poor osseointegration, and even implant failure [2]. To overcome these risks, computer-assisted implant surgery (CAIS) was developed, comprising static (s-CAIS) and dynamic (d-CAIS) systems that assist clinicians in executing preoperative plans with improved control and safety [2].

Despite benefits, each of these technologies has limitations. Static guides, while physically stabilizing the drill path, are often inflexible, prone to intraoperative fractures or displacement, and restricted by patient-specific anatomical challenges, especially in posterior regions [3]. On the other hand, dynamic navigation systems offer real-time visual tracking and adaptive guidance, but they require the clinician to divide attention between the screen and surgical site making these machines dependent on user experience [3]. Moreover, all procedures, regardless of the technology used, still rely on manual execution, introducing the possibility of errors from human errors [4].

Robotic computer-assisted implant surgery (r-CAIS) addresses these shortcomings by combining the physical stability of static systems with the real-time adaptability of dynamic systems. Robotic platforms have demonstrated the ability to plan, guide, and even autonomously perform implant site preparation and placement with high accuracy [5]. Based on their level of autonomy, robotic systems can be categorized as active (fully autonomous), semi-active (clinician-guided), or passive (clinician-controlled) [6].

Similarly, the field of prosthodontics has also embraced digital workflows to increase the efficiency and accuracy of fixed and removable restorations. Technologies such as CAD/CAM have enabled the fabrication of crowns, bridges, and dentures with improved precision and reduced clinical and laboratory time [7]. However, despite these improvements, key steps in prosthodontic treatment such as tooth preparation and denture tooth arrangement still rely heavily on operator skill and manual techniques, which can introduce variability, errors, and the need for retreatment [8].

Robotic systems in prosthodontics have been developed to automate some dental procedures involving tooth arrangements and crown preparations. For example, robotic arms programmed with digital reduction parameters can perform highly standardized and accurate tooth preparations for crowns and veneers, improving consistency and minimizing iatrogenic damage [9]. Similarly, robots have been applied in denture tooth arrangement, replacing labor-intensive manual setups and reducing dependency on the technician’s expertise [8]. These robotic interventions not only reduce chairside time and operator fatigue but also improve patient comfort and treatment predictability.

Although still in early developmental stages within prosthodontics, robotic applications are showing promise in transforming how dental restorations are designed and delivered [10]. Their use has been associated with higher precision, better adaptation of prosthetic margins, fewer complications, and reduced need for adjustments outcomes that mirror those achieved in robotic-guided implant placement [2]. Reviews published in literature have explained the importance of robotics in relation to general dentistry [1, 4, 5, 7, 11, 12]. However, none have explained the importance and application of robotics in prosthetic and implant dentistry together.

Moreover, the integration of prosthetic and implant dentistry in evaluating robotic technology was the direct impact both the procedures have in patient related outcomes. Studies have reported that with ill fitted prosthesis it is difficult to place an implant and vice versa [5, 7]. Assessing robotics across this combined continuum allowed for a more meaningful evaluation of how these technologies reduce complications, improve accuracy, and ultimately enhance treatment predictability and patient satisfaction throughout the entire rehabilitation process.

Hence, a comprehensive understanding of current applications, limitations, and emerging innovations is essential to inform future research, clinical guidelines, and policymaking. This scoping review aims to synthesize the current state of knowledge on the utilization of robotics in prosthetic and implant dentistry, identify research gaps, and provide insights into the pathway toward clinical translation and broader adoption. It focused on the ability of robotics to overcome limitations of conventional methods, evaluate the stages of robotic developments, autonomy classification, and determine benefits of robotics in prosthetic and implant dentistry.

Materials and methods

The current scoping review followed a systematic method to evaluate the current status of robotics in prosthetic and implant dentistry. This review was adhered to the guidelines developed by Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA-ScR) [13]. The initial search was made on PubMed following the Mesh terminologies. Limitations on language and year of publication were made, and the search was bounded from January 2015 till May 2025. The search results were imported into Endnote20 to remove duplication. The review question, Population, Content, and Context framework, inclusion and exclusion criteria, search terms, databases, and keywords are presented in Table 1.

Table 1.

Review question, PCC framework, inclusion and exclusion criteria, databases searched, and Mesh terminologies and manual search sources

Items	Description
Research Question	What are the current applications, effectiveness, and limitations of robotics in prosthetic and implant dentistry?
PCC Framework	Population (P): Individuals requiring prosthodontic or implant-related dental care. Concept (C): Use of robotic systems in prosthetic and implant dentistry, including their applications, clinical effectiveness, technological advancements, and limitations. Context (C): Clinical and laboratory settings where robotics are employed in prosthetic and implant procedures.
Eligibility Criteria	Inclusion Criteria: In vitro studies, clinical trials, case series, and observational studies. Studies focused specifically on robotic applications in prosthetic and/or implant dentistry. Exclusion Criteria: Studies focused solely on robotic applications in other dental specialties. Review articles, editorials, expert opinions, and conference abstracts without complete data
Databases Searched	PubMed, SCOPUS, EMBASE, Web of Science, Google Scholar
Search Terms / Keywords	(“Prosthodontics“[MeSH Terms] OR “prosthodontic“[All Fields] OR “prosthetic dentistry“[All Fields] OR “dental prosthes*“[All Fields] OR “removable prosthodontic*“[All Fields] OR “fixed prosthodontic“[All Fields] OR “implant dentistry“[All Fields] OR “dental implant*“[All Fields]) AND (“robot*“[All Fields] OR “robotics“[MeSH Terms] OR “robotic technology“[All Fields] OR “robotic dentistry“[All Fields] OR “robotization“[All Fields] OR “robotized“[All Fields] OR “automation“[All Fields]) (Supplementary Table 1)
Additional Search Sources	Manual searching of reference lists from selected studies, OpenGrey

Two reviewers (H.L. and A.J.) screened the titles and the abstracts of the extracted articles using a predefined inclusion criterion. If abstract was not clear authors read full text article. After considering articles for potential relevance the reviewers reached a consensus on those to include and if there was any disagreement it was resolved by discussion with third investigator (A.P.). Three pre-designed data collection forms were used to collect data from the included studies (1) Prosthetic dentistry; (2) Implant dentistry; (3) Developmental stage of robots in prosthetic and implant dentistry including, author details, study design, robotics used, developmental levels of robotics, type of surgery, implant placement method, comparison if required, key findings and level of autonomy. Additional information included prosthetic and implant procedures in which robots were used and the results.

Results

A total of 533 studies on the use of robotic technology in prosthetic and implant dentistry were identified by different database search, and five additional studies were extracted through manual searching. Complete screening procedure led to a full text selection of 41 articles, from which seven studies were extracted for inclusion in prosthetic dentistry and fourteen studies in implant dentistry (Fig. 1). The summary of characteristics included studies focusing on application of robotics in prosthetic and implant dentistry are presented in Tables 1, 2 and 3.

Fig. 1.

PRISMA flowchart 2020

Table 2.

Characteristics of studies on the application of robotics in prosthetic dentistry

Authors / Year / Country	Study Design	Sample Population / Size	Prosthetic Procedure	Type of Robotics Used	Intervention Details	Comparator	Outcomes Measured	Key Findings	Technologies Integrated	Limitations
Sun Jian-Peng et al. 2023, China [14]	In vitro proof-of-concept study	STL model of maxillary central incisor (plaster teeth)	Veneer preparation (fenestration type)	Dobot Magician serial joint manipulator (0.2 mm repeat positioning accuracy)	- Digital interactive design system for veneer preparation. - Robot-assisted trajectory planning using B-spline curves and NURBS surfaces. - Layered preparation (0.15 mm, 0.3 mm, 0.45 mm).	Manual preparation (implied by traditional methods)	Preparation depth errors (X, Y, Z directions). - Surface flatness errors. - System error confidence intervals.	- Maximum errors : 0.180 mm (X), 0.104 mm (Y), 0.142 mm (Z). - Confidence interval stable within 0.30 mm. - Feasibility of digital design and robot-assisted preparation confirmed.	Mesh segmentation. - B-spline/NURBS interpolation. - OpenGL/C + + for software. − 3D printing for tooth models.	-Plaster teeth used (hardness difference vs. real teeth). - Robot accuracy limited (0.2 mm). - Algorithm not universal (tooth-specific). - No clinical validation -Experimental model lacks biological variability; higher-precision robots suggested for future work.
Jiang Jin-Gang et al. 2013, China [8]	Experimental study (proof-of-concept)	1 edentulous patient (64-year-old male)	Full denture manufacturing (dental arch generation)	Multi-manipulator tooth-arrangement robot with dental arch generator (8 stepper motors)	- Power function model for dental arch generation. Synchronized control of 8 stepper motors. - High-resolution timing for pulse control. -Multi-task scheduling algorithm.	Traditional manual dental arch generation	- Single-point error (max: 1.83 mm). - Repeated positioning accuracy (0.09 mm). - Trajectory matching (theoretical vs. actual).	- Automated dental arch generation feasible. - High precision (0.09 mm repeatability). - Matched patient-specific arch parameters. -Smooth tooth arrangement achieved.	- Power function model (α, β parameters). - Industrial PC + PC6401 control card. - multi-threading for synchronization. - 3D laser scanning for jaw measurements.	- Single patient case. - Limited to mandibular dentition. - High computational dependency. -No long-term clinical validation.
Jiang Jin-Gang et al. 2022, China [9]	Experimental study	N/A	Full crown adjacent surface tooth preparation	Dobot Magician robotic arm (4-DOF, 0.2 mm repeatability)	- NURBS-based trajectory planning. - Augmented Reality (AR) mobile app (DPAS AR) for visualization and control. - Bluetooth-enabled robot control.	Traditional manual preparation	- Relative fixed-point errors (X, Y, Z axes). - Relative standard deviation (RSD). Confidence intervals.	- Max errors : 0.24 mm (X), 0.29 mm (Y), 0.37 mm (Z). - RSD < 11% in all axes. - AR improved visualization and human-robot interaction.	- NURBS curve interpolation. - ARCore SDK (Unity3D). - Bluetooth GATT protocol. − 3D model preprocessing (Geomagic/Creo).	- No clinical validation (in vitro only). - Limited to posterior teeth. - Dependency on high-precision robotic hardware.
Yuan, F. et al. 2021, China [15]	Experimental	30 resin teeth (molars) divided into 6 groups	Automatic tooth preparation for full metal crowns	Robotic system with femtosecond laser (TANGERINE, Amplitude Systems)	Light-off delay settings varied from 180 ms to 480 ms (40 ms intervals) during laser ablation	Different light-off delay groups	Taper of tooth preparation; average distance between scanned and predesigned data	- Mean taper decreased with increased light-off delay (39.268° to 25.393°).	- Femtosecond laser (1030 nm, < 300 fs pulse width, 200 kHz repetition rate). - CAD/CAM software. − 3D scanning and Geomagic Studio for analysis	- Used resin teeth, not natural teeth. - Clinical relevance of taper values may differ. - Carbonization observed at 480 ms delay.
Fusong Yuan & Peijun Lyu, 2020, China [16]	Preliminary experimental study	20 extracted human first molars (15 used for full-crown preparation, 5 for single-depth ablation measurement)	Full-crown tooth preparation	Robot-controlled ultrashort pulse laser system (6-degree-of-freedom laser beam arm, intraoral working end, tooth positioner)	Automated tooth preparation using a femtosecond laser (1064 nm wavelength, 15 ps pulse width, 100 kHz repetition frequency, 30 W power). Layer-by-layer ablation with CAD/CAM software for path planning.	N/A	- Single depth of ablation - Occlusal preparation depth - Convergence angle - Overall preparation error -Time efficiency	- Single depth of ablation: 45 ± 3.55 μm - Occlusal depth error: 0.097 ± 0.022 mm (target: 2 mm) Convergence angle error: 0.98 ± 0.35° (target: 6°) - Overall mean error: 0.05–0.17 mm - Average preparation time: 17 ± 1.77 min	- CAD/CAM software - Ultrashort pulse laser - Intraoral 3D scanner (TRIOS) − 6-DOF robotic arm - Geomagic software for 3D analysis	- Limited to shoulder-less full-crown preparations. - No clinical trials (in vitro only). - Light blocking issues in stripping method not resolved. The robot-controlled laser system demonstrated high precision and feasibility for automated tooth preparation. - Further research needed for complex preparation, veneers, inlays).
Fusong Yuan et al. 2016, China [18]	Experimental	10 extracted mandibular first molars	Automated tooth preparation (full crown)	Robot-controlled femtosecond laser system (6-DOF arm, confined space setup)	- Laser: Yb: KYW femtosecond laser (1025 nm, < 400 fs pulse width, 100 kHz repetition rate, 4.4 W power, 500 mm/s scan rate). - Cooling: Air cooling vs. no air cooling. -Setup: Thermocouples in pulp cavity and ambient environment.	Air cooling vs. no air cooling	Temperature in pulp cavity (°C) - Ambient temperature (°C) - Statistical significance (p-value)	- Pulp cavity temperature: - With air cooling: 32.43 ± 1.56 °C - Without air cooling: 52.31 ± 1.65 °C (p < 0.05). - Ambient temperature: - With air cooling: 22.98 ± 1.88 °C - Without air cooling: 28.53 ± 1.53 °C (p < 0.05).	- Femtosecond laser (JenLas® D2.fs) - Thermocouples + data logger - CAD/CAM (Geomagic Studio) - Robotic arm for precision ablation	- In vitro study (no clinical validation). - Long ablation time (~ 22 min). - No simultaneous multi-point temperature measurement. -Air cooling effectively reduces heat generation, keeping pulp temperatures below damage thresholds (5.6 °C increase).
Fusong Yuan et al. 2017, China [17]	Preliminary experimental study	15 freshly extracted human intact first molars	Full crown tooth preparation	Intraoral automatic laser-controlled micro-preparation unit with 6-DOF light guiding arm and high-speed galvanometers	- Robotic device controlled an ultra-short pulse laser (USPL) beam (wavelength 1,064 nm, pulse width 15 ps, output power 30 W, repeat frequency rate 100 kHz). - Tooth preparation performed via 3D motion planning software and layered preparation method. - Preparation time recorded, and accuracy evaluated using Geomagic Studio and Image ware software.	N/A	- Overall shape error (mm) - Occlusal reduction in depth error (mm) Convergence angle error (°) - Preparation time (minutes)	- Average shape error: 0.05–0.17 mm. - Occlusal reduction in depth error: ~0.097 mm. Convergence angle error: ~1.0°. - Average preparation time: 17 min. -Results validated the feasibility and accuracy of the automatic tooth preparation technique.	- Ultra-short pulse laser (USPL) − 3D motion planning software - CAD/CAM software - Intraoral 3D scanner - Robotic system with high-speed galvanometers	-Limited to non-shouldered full crown preparations. - No direct comparison with manual preparation. - Preparation efficiency could be improved (volume preparation method). - Safety mechanisms (real-time monitoring) require further study. - Study conducted in a phantom head, not in vivo.

N/A not applicable, CAD/CAM computer-aided design and computer-aided manufacturing

Table 3.

Characteristics of studies on the application of robotics in implant dentistry

Authors / Year / Country	Study Design	Sample Population / Size	Type of Robotics Used	Application Area	Intervention Details	Comparator	Outcomes Measured	Key Findings	Technology Used	Limitations
Zonghe Xu et al. 2023, China [30]	In vitro study	108 mandibular models (216 implants)	- Active Robot (ARG) - Passive Robot (PRG) - Semi-Active Robot (SRG) - Active Dynamic Navigation (ADG) - Passive Dynamic Navigation (PDG)	Dental implant placement	- Each group placed 2 implants (positions 31 and 36) 12 times. - Surgeons with varying experience levels performed the procedures.	Comparison between robotic systems (ARG, PRG, SRG) and dynamic navigation systems (ADG, PDG)	- Coronal deviation (mm) - Apical deviation (mm) - Axial deviation (°)	- ARG had the lowest deviations (coronal: 0.29 ± 0.15 mm; apical: 0.29 ± 0.15 mm; axial: 0.61 ± 0.25°). - Number of surgeries did not affect accuracy (P > 0.05).	- Dynamic navigation systems: Optical trackers (active/passive infrared). - Robotic systems: UR5 robotic arm, automatic path planning. - CBCT (i-CAT FLX V10) for imaging.	- In vitro study limits clinical applicability. - Homogeneous resin models lack real bone density variations. .
Shasha Jia et al. 2025, China [19, 20]	Retrospective study	60 implants in 39 patients	autonomous dental implant robotic (ADIR)	Implant placed in mandible	20 implants were placed with ADIR system in mandibular premolar and molar region. 19 implants were placed according to computer assisted dynamic system.	compare the accuracy of implants placed with an ADIR system with those placed with sCAIS	The coronal, apical, and angular deviations were measured and analyzed	The mean ±standard deviation of the ADIR system group and sCAIS group were 0.43 ± 0.18 mm versus 1.31 ± 0.62 mm (P<0.001), 0.56 ± 0.18 mm versus 1.47 ± 0.65 mm (P<0.001), and 1.48 ± 0.59 degrees versus 2.42 ± 1.55 degrees (P=0.003),	CBCT for imaging. Dynamic navigation systems: Optical trackers	The accuracy of ADIR system was significantly higher than aCAIS system.
Chen et al. 2023 China [21]	Single blinded RCT	31 implants in 28 patients	Robotic system	Implant placement in partially edentulous patients	12 implants were placed using a robotic system with the help of trained operator and post operatively OPG was taken for analysis.	The deviations were compared with objective performance goals (OPGs) from reported studies of fully guided static computer-assisted implant surgery (CAIS) and dynamic CAIS	multiple linear regression analysis	mean angle deviation of 2.81 ± 1.13°	OPG machine, robotic system, CAIS system	The robotic system was more accurate in placing the implant than CAIS.
Thu Ya Linn et al. 2023, China [22]	Experimental study	67 drilling sites on Phantoms	Robotic system	Drilling sites on Phantoms with three different implant sizes (Ø = 3.5 × 10 mm, 4.0 × 10 mm, 5.0 × 10 mm)	robotic procedure was performed using software for calibration and step-by-step drilling processes. After robotic drilling, deviations in the implant position from the planned position were determined.	None	The angulation, depth, and coronal and apical diameters on the sagittal plane of sockets created by human and robotic drilling were measured.	The deviation of the robotic system was 3.78° ± 1.97° (angulation), 0.58 ± 0.36 mm (entry point), and 0.99 ± 0.56 mm (apical point)	Robotic system	Robotic surgical systems offer high accuracy and reliability for planning and placing small implants.
Bolding and Rebyee 2022, USA [23]	prospective single-arm clinical trial	38 endosteal implants in 5 participants	haptic robotic guidance	Three dual-arch and 2 single-arch	A virtual preoperative plan was made and aligned to the surgical site using a bone-mounted fiducial splint created from a CBCT scan. During surgery, implants were placed according to the virtual plan with the aid of a robotic guidance arm and haptic constraints.	Traditional implant surgery	Mean and standard deviation, angulation before and after placement of implants with robotic arms.	The mean ±standard deviation global angular deviation was 2.56 ± 1.48 degrees, while the crown of the placed implant showed a deviation from the plan of 1.04 ± 0.70 mm and the apex of 0.95 ± 0.73 mm. The signed depth deviation was 0.42 ± 0.46 mm	CBCT and haptic robotic arms	Haptic robotic arms are efficient to place implant at any edentulous site. However, more clinical studies are required.
Baoxin Tao et al. 2022, China [24]	Experimental study	480 dental implants were placed in 80 phantoms	Hybrid Robotic System for Dental Implant Surgery, HRS-DIS	Implants placed on phantoms	The entry, exit and angle deviations of the implants in 3D world were measured after pre-operative plans and postoperative cone-beam computed tomography (CBCT) fusion.	dynamic navigation system (Beidou-SNS)	linear mixed model with a random intercept	mean entry deviation of 0.96 ± 0.57 mm vs. 0.83 ± 0.55 mm (p = 0.04), a mean exit deviation of 1.06 ± 0.59 mm vs. 0.91 ± 0.56 mm (p = 0.04), and a mean angle deviation of 2.41 ± 1.42° vs. 1 ± 0.48° (p < 0.00).	CBCT, HRS-DIS, and Beidou-SNS	The robotic system achieved greater implant positioning accuracy than the dynamic navigation system, indicating that the HRS-DIS prototype may be a valuable tool in dental implant surgery.
Yang et al. 2024, China [25]	randomized controlled clinical trial	140 patients (70 in RAIS group, 70 in FHIS group)	Semi-active robotic system (THETA, Hangzhou Jianjia Medical Technology Co., Ltd)	Single dental implant placement	Robot-assisted implant surgery (RAIS) with semi-active robotic system	Free-hand implant surgery (FHIS)	Platform, apex, and angular deviations; safety; surgical morbidity	RAIS showed superior accuracy (platform: 0.76 ± 0.36 mm, apex: 0.85 ± 0.48 mm, angle: 2.05 ± 1.33°) compared to FHIS (platform: 1.48 ± 0.93 mm, apex: 2.14 ± 1.25 mm, angle: 7.36 ± 4.67°). No significant safety issues.	THETA robotic system, CBCT, 3D Slicer software	Limited to single-quadrant implant placement; preoperative preparation time increased; no comparison with dynamic navigation or static guides.
Li et al. 2023, China [26]	Preliminary research (case study + in vitro experiment)	1 clinical case (2 zygomatic implants) + in vitro model (10 zygomatic implants, 20 alveolar implants)	Autonomous robotic system (Remebot, Beijing Rui Yi Bo Technology Co., Ltd.)	Zygomatic implant placement and immediate full-arch prosthesis	Robotic-assisted zygomatic implant placement with customized drills and markers	In vitro comparison with alveolar implants	Entry point error, exit point error, angle error	Zygomatic implants: entry error 0.78 ± 0.34 mm, exit error 0.80 ± 0.25 mm, angle error 1.33 ± 0.41°. Comparable to alveolar implants (p > 0.05). Clinical cases showed similar accuracy.	Remebot robotic system, CBCT, 3D printing, customized drills/markers	Limited sample size (1 clinical case); challenges with mouth opening for long drills; requires further clinical validation.
Chen et al. 2023, China [27]	In vitro pilot study	10 partially edentulous models (20 implant sites)	Semi-active robotic system (THETA, Hangzhou Jianjia Robot Co.)	Dental implant placement	Robotic-assisted implant surgery with THETA system	Dynamic navigation system (Yizhimei)	Platform, apex, and angular deviations	THETA group: platform 0.58 ± 0.31 mm, apex 0.69 ± 0.28 mm, angle 1.08 ± 0.66°. Yizhimei group: platform 0.73 ± 0.20 mm, apex 0.86 ± 0.33 mm, angle 2.32 ± 0.71°. THETA showed superior angular accuracy (p < 0.001).	THETA robotic system, CBCT, 3D Slicer software	Small sample size; in vitro study lacks clinical variability; no long-term follow-up.
Qiao et al., 2023, China [28]	Translational study (in vitro + clinical case series)	3D printed resin models (12 holes each in 3 groups) Clinical: 21 patients, 28 implants	Semi-autonomous collaborative robot (Cobot; Langyue dental surgery robot) with haptic feedback & machine vision	Dental implant placement	Robotic drilling with/without lock-on structure in sterilized/registration handpieces; Clinical: Cobot-assisted implant placement using customized patient splint, CBCT planning, real-time haptic and visual feedback, validated accuracy before drilling	Conventional freehand and guided approaches (from literature)	Platform deviation, apex deviation, angular deviation; surgery time; bone thickness; complications	Cobot-assisted placement showed high accuracy: mean platform deviation ~ 0.53 mm, apex deviation ~ 0.56 mm, and angular deviation ~ 0.79°. Superior to reported accuracy of static and dynamic navigation.	- Cobot (Universal Robots UR3, Denmark) - Micro infrared camera (≤ 5 cm tracking) - CBCT (Planmeca Viso G7) - Planning software (Naviguide) - ATI haptic sensor - Lock-on structure for handpiece stability	- Not flexible enough in posterior region with limited space - Accuracy depends on CBCT quality - Risk of drill sideslipping - Robotic movement requires re-positioning time - Generalizability to more complex clinical scenarios remains to be tested.
Ping Li et al., 2023, China [29]	Retrospective case series	10 patients (8 males, 2 females); 59 implants in fully edentulous or terminal dentition patients	Autonomous robotic system (Remebot)	Full-arch implant placement in fully edentulous patients	Implant placement planned using CBCT and coDiagnostiX software; customized positioning marker fabricated and fixed; intraoperative registration and calibration; implant osteotomy and placement performed by autonomous robotic arm; accuracy measured via pre- and postoperative CBCT alignment	No direct control or comparison group	Global coronal deviation, global apical deviation, angular deviation	Autonomous robotic surgery demonstrated high accuracy: global coronal deviation 0.67 ± 0.37 mm, apical 0.69 ± 0.37 mm, angular 1.27° ± 0.59°; no surgical complications reported	Remebot robotic system (Beijing Baihui Weikang Tech) - Optical tracker and robot arm - CBCT (NewTom VGI) - Planning software: coDiagnostiX - Marker design: Exocad Printing system: Ultracraft (HeyGears) with surgical guide resin	- No comparative cohort (no s-CAIS or d-CAIS) - Monocentric and small sample size - Some implants placed freehand due to high torque - Time-consuming workflow with multiple CBCT scans - High cost and steep learning curve
Xu et al., 2023, China [30]	In vitro comparative study	30 mandibular phantoms (60 implants total)	• Semi-active (Remebot, SR) • Active (Yekebot, AR) • Passive (DentRobot, PR)	Dental implant placement	Each robot placed 2 implants (#31 and #36) on 10 phantoms. Pre-op CBCT planning and robot-specific calibration protocols. Robots differed in the level of human-robot interaction from fully autonomous (AR) to human-guided (PR). Implant surgery was performed, followed by post-op CBCT for accuracy measurement.	Comparison between SR, PR, and AR robots	• Preparation time • Operation time • Coronal deviation • Apical deviation • Axial deviation	• PR had shortest operation time but lowest accuracy • AR had highest preparation time but best accuracy • SR had balance of efficiency and accuracy • Human-robot interaction significantly affects outcomes	• Remebot (SR), Yekebot (AR), DentRobot (PR) • Optical trackers: Micron Tracker, Fusion Track, Polaris Vicra • Navigation Software: RemeDent, DentalNavi, Dcarer software • Implant planning via CBCT and intraoral scans • Implant system: Nobel PMC • Robotic platform: UR5 (Universal Robots)	• In vitro setting only • Did not test edentulous jaws • Accuracy may differ in real clinical conditions • Differences in registration algorithms between robots • Results need validation in animal or clinical models for generalization
Shi et al. 2024, China [31]	Pilot randomized controlled trial	20 patients (10 per group)	Theta dental implant robot system (collaborative robotic arm with haptic feedback and machine vision)	Single-site dental implant placement	Robot-assisted implant placement with real-time feedback and haptic constraints	Freehand implant placement	Positional accuracy (platform/apex global deviation, angular deviation), surgical morbidity, complications, bone thickness, patient-reported outcomes	Robot-assisted placement showed significantly better positional accuracy (p < 0.05 for most measures) with comparable safety and morbidity.	Theta robot (Jianjia Ltd.), DTX Studio™ Implant software, CBCT imaging, 3D Slicer software	Small sample size, one case excluded due to calibration error; limited to single implants.
Tahir et al. (2019), Italy [32]	Experimental study	Not specified (prototype testing)	Robotic mastication simulator with Stewart PKM	Dental implant testing and prosthodontics	Development of a robotic mastication simulator to replicate human mastication forces	None	Force replication, load distribution, hydraulic system performance, sensor accuracy	PKM effectively replicated mastication forces (200–2000 N range). Hydraulic system validated. Sensor recorded interactive loads accurately.	Stewart PKM, hydraulic actuators, capacitive force sensors, load-sensing element	No human subjects tested, prototype stage. Limitations include impulsive displacements during operation.

CBCT cone beam computed tomography, aCAIS static computer-assisted implant surgery, dCAIS dynamic computer-assisted implant surgery, AR active robot, SR semi-active robot, PR, passive robot

Prosthetic dentistry

Seven experimental studies conducted between 2013 and 2023 in China were included, highlighting various applications of robotics in prosthetic dentistry [8, 9, 14–18]. These studies explored robot-assisted procedures such as veneer preparation, full crown preparations, and automated denture arch generation, primarily in non-clinical settings using extracted teeth, resin models, or STL-reconstructed dental arches (Table 2).

Sun et al. developed a digital design and trajectory planning system using a Dobot Magician robotic arm for veneer preparation [14]. Their findings demonstrated clinically acceptable precision, with maximum preparation depth errors ranging from 0.104 mm to 0.180 mm across spatial axes [14].

Yuan et al. [18] and Jiang et al. [9] explored the use of robot-controlled femtosecond laser systems and mechanical arms in preparation of full crown. These systems achieved remarkable precision, with mean preparation errors generally remaining within 0.05–0.17 mm, and occlusal depth deviations consistently under 0.1 mm. Yuan et al. reported convergence angle errors under 1°, validating the accuracy of layered, robot-guided preparation techniques [18]. A novel integration of augmented reality (AR) with robotic preparation was demonstrated by Jiang et al., by employing an AR-enhanced planning app to guide robot-assisted full crown preparation [9]. The AR visualization improved interaction and control, and the robot achieved preparation errors under 0.4 mm with relative standard deviations below 11%.

Jiang et al. designed a multi-manipulator robotic system with eight stepper motors for automated dental arch generation for complete denture fabrication [8]. The robot effectively replicated patient-specific arch parameters, achieving repeated positioning accuracy of 0.09 mm [8]. Two studies focusing on laser-based robotic preparation revealed that femtosecond laser systems offer high-precision ablation with minimal heat damage when appropriate cooling is applied [15, 18]. Fig. 2 explains the workflow of robots in prosthetic dentistry.

Fig. 2.

Workflow of Robot in crown preparation. A complete preparation model is generated through mesh fusion techniques, integrating all modified regions into a unified structure. The processed model is exported as an output for robotic trajectory planning, enabling automated and precise tooth preparation procedures

Table 3 explained the autonomy and level of development of robotics in prosthetic dentistry. According to the included studies, five studies utilized active robotics with stage II and III development phase [9, 14, 15, 17, 18]; whereas none of the robots used in prosthetic dentistry were passive and stage IV of development to be utilized in clinics.

Implant dentistry

A total of fourteen studies were included in demonstrating robotic systems in dental implant placement compared to traditional methods [20, 22, 26, 29], static guides [24, 28, 31, 33, 34], and dynamic navigation systems [19, 25, 28, 32] (Table 3). Xu et al. conducted an extensive in vitro study comparing three robotic groups (active, semi-active, passive) and two dynamic navigation systems, finding significantly lower coronal, apical, and angular deviations in robotic systems [30]. Similarly, Jia et al. reported that autonomous dental implant robots achieved markedly better precision than static computer-assisted implant surgery (s-CAIS), with coronal and apical deviations [20].

Chen et al., in a single-blinded RCT, showed robotic systems were more accurate than fully guided CAIS, although the difference was not statistically significant [27]. Linn et al. demonstrated robotic drilling on phantom models that yielded deviation metrics within clinically acceptable ranges [22]. In a clinical trial by Bolding and Reebye, haptic robotic guidance achieved a mean angular deviation of 2.56°, and coronal/apical deviations under 1 mm [23].

Tao et al. compared a hybrid robotic system (HRS-DIS) with dynamic navigation (Beidou-SNS), finding robotic implants had significantly better angular accuracy (1.0° vs. 2.41°, p < 0.001) [24]. Yang et al., through a randomized clinical trial, showed the semi-active robotic system (THETA) significantly outperformed freehand surgery in all accuracy parameters [25]. Similarly, in a pilot study, Chen et al. confirmed THETA’s superior angular precision over dynamic navigation in partially edentulous models [27].

Li et al. reported that robotic-assisted placement of zygomatic implants demonstrated comparable accuracy to alveolar implants, despite anatomical complexity [26]. Qiao et al. validated the precision of a collaborative cobot system, achieving deviations below 0.8° better than static and dynamic guided methods [28]. Ping Li et al. supported these findings in a clinical case series involving fully edentulous patients, noting global deviations under 0.7 mm and 1.3°, with no surgical complications [26].

Xu et al. evaluated human-robot interaction levels (active, semi-active, passive), showing that while passive robots had faster operation times, their accuracy was inferior; active robots had the best accuracy but required longer setup [19]. Shi et al. supported this, showing the Theta system significantly outperformed clinical placement in a randomized trial [31]. Tahir et al. developed a robotic mastication simulator, validating its technical performance [32]. The workflow of robots in implant dentistry is explained in Fig. 3.

Fig. 3.

Workflow of robotics in dental implant surgery. This workflow demonstrates the integration of imaging, digital planning, and robotic execution to enhance accuracy, reproducibility, and clinical outcomes in implant dentistry

Of included 14 studies, robots included in11 studies had semi-active autonomy levels [20–22, 25–29, 31], two were at passive level [23, 30] and one at active level [31] (Table 4). Ten robotic studies included in implant dentistry utilized robots at stage IV of development while three studies were at stage II development. Fig. 4 demonstrated the studies included in the current reviews and which prosthetic and implant procedures are performed.

Table 4.

Autonomy and development stage of robots in prosthetic and implant dentistry

Study ID	Development stage	Autonomy levels	Description
Prosthetic Dentistry
Sun Jian-Peng et al. (2023), China [14]	III	Semi-active	The Dobot Magician robotic arm assists with tooth preparation, independently preparing the implant bed and installation. Operator guidance is needed for positioning the arm in and out of the mouth.
Jiang Jin-Gang et al. (2013), China [8]	II	Active	A fully automated tabletop robotic arm equipped with a high-speed handpiece is designed for tooth preparation. It autonomously navigates in and out of the oral cavity, performs implant bed preparation, and carries out implant placement. The operator is solely responsible for changing drills, providing commands, and overseeing the procedure.
Jiang Jin-Gang et al. (2022), China [9]	III	Active	An automated laser ablation system designed for tooth crown preparation is capable of independently accessing and existing the oral cavity, preparing the implant site, and placing the implant. The clinician’s role is limited to changing the drill, providing operational commands, and supervising the process.
Yuan, F. et al. (2021), China [15]	III	Active	Resin teeth were mounted on a jaw model. The robotic system is capable of autonomously entering and exiting the oral cavity, performing implant bed preparation, and inserting the implant. The operator’s role is limited to changing the drill, giving instructions, and overseeing the procedure.
Fusong Yuan & Peijun Lyu (2020) China [15]	II	Active	Freshly extracted mandibular molars were secured in a dental model. The robotic system autonomously enters and exits the oral cavity, prepares the implant site, and performs implant placement. The operator’s duties are limited to changing drills, issuing commands, and supervising the procedure.
Fusong Yuan et al. (2017) China [18]	III	Semi-Active	A Yb: KYW diode-pumped, solid-state, thin-disk, femtosecond laser controlled by a robotic device.
Fusong Yuan et al. (2016) China [17]	III	Active	An intraoral automated micro-preparation unit utilizing an ultra-short pulse laser (USPL) beam for precise laser control is designed for dental procedures. It can autonomously access and exit the oral cavity, carry out implant bed preparation, and place the implant. The operator is only required to replace the drill, provide instructions, and monitor the system’s operation.
Implant Dentistry
Zonghe Xu et al. (2023), China [19]	II	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Shasha Jia et al. 2024, China [20]	IV	Semi active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Chen et al. 2023 China [27]	IV	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Thu Ya Linn et al. 2023, China [19, 22]	II	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Bolding and Rebyee 2022, USA [23]	IV	Passive	This robot cannot act independently; the operators must guide the robotic arms to enter and exit the mouth, prepare the implant bed, and put the implant.
Baoxin Tao et al. 2022, China [24]	II	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Yang et al. (2024), China [25]	IV	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Li et al. (2023), China [29]	IV	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Chen et al. (2023), China [21]	IV	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Qiao et al., (2023), China [28]	III	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Ping Li et al., 2023, China [26]	IV	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.
Xu et al., 2023, China [30]	IV	Semi-Active (Remebot), active (Yekebot), and passive (DentRobot)	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth. This robot can independently enter and exit the mouth, prepare the implant bed, and insert the implant. The operator’s only responsibility is replacing the drill, issuing instructions, and monitoring the robot’s operation. This robot cannot act independently; the operators must guide the robotic arms to enter and exit the mouth, prepare the implant bed, and put the implant.
Shi et al. (2024) China [31]	III	Active	This robot can independently enter and exit the mouth, prepare the implant bed, and insert the implant. The operator’s only responsibility is replacing the drill, issuing instructions, and monitoring the robot’s operation
Tahir et al. 2019, Italy [32]	II	Semi-active	The robot prepares the implant bed and installs it on its own, but an operator must guide the robotic arm in and out of the mouth.

I. Research level: The robot is under development, which includes design and analysis of its hardware or software components

II. Phantom experiments level: Initial development of the robot is complete and experiments along with corresponding improvements are being conducted on phantoms

III. Clinical validation level: The robot has been developed and is undergoing clinical trials but has not yet received national medical approval or entered the market

IV. Clinical application level: The robot is available on the market and is currently used in the treatment of dental diseases

Fig. 4.

Studies reporting practical application of robotics in the implant and prosthetic dentistry

Discussion

This scoping review offers a comprehensive evaluation of the current status of robotic applications in both implant dentistry and prosthodontics, highlighting their impact on clinical efficiency, accuracy, and procedural outcomes. While robotics in both fields have proven effective in minimizing human error and improving treatment predictability, the extent of time-saving benefits appears to be application-dependent. These findings suggest that robotic technology holds strong potential to transform both implant and prosthetic dental practices, though further refinement and integration into clinical workflows are needed to optimize efficiency across all applications.

Robotics in prosthetic dentistry

In 2001, Peijun et al. [35] and Zhang et al. [36] conducted early robotics experiments in prosthodontics, which encountered major technical issues such as slow 3D rendering, occlusal inaccuracies, unstable components, and inadequate polymerization power. Challenges in coordinate control and timing precision hindered broader clinical translation, and research in this area has since stagnated, potentially due to competition from digital technologies like CAD-CAM and 3D printing.

Robotic arms equipped with high-speed cutting tools, as demonstrated by Otani et al. [37] and Jiang et al., [9] achieved preparation accuracy comparable to manual techniques. However, limitations included inadequate precision, inconsistent depth control, and variability in velocity profiles. Software enhancements such as Unity3D-based trajectory planning and augmented reality have improved operator interaction but did not fully resolve mechanical limitations [38].

Laser-based robotic systems, particularly those using picosecond or femtosecond photoablation, have shown promise in achieving clinically acceptable crown preparations [15]. Despite early challenges with heat control, convergence angles, and laser output, later studies by Yuan et al. developed optimized protocols, demonstrating improved accuracy and preparation quality [15, 16]. Components like multi-axis robotic manipulators, digital scanners, and end-effectors contributed to enhanced control and fit. However, preparation times remained longer than traditional methods due to factors such as scanning, planning, and calibration steps. Studies reported preparation durations ranging from 17 min per tooth to several hours, depending on the substrate and technique [39, 40].

The challenges include the absence of real-time recalibration mechanisms, lack of temperature monitoring during ablation, limited safety protocols, and energy losses in laser delivery systems. In addition, the large size, high cost, and maintenance demands of current robotic systems may limit their integration into routine practice [4]. The need for modified control systems, enhanced thermal management, and streamlined workflows is required. AI can augment robotic systems by enabling adaptive behavior, real-time decision-making, and enhanced environmental awareness [3]. Continued innovation, along with robust clinical validation, will be essential to address current limitations and support the safe and effective adoption of robotic systems in prosthodontics.

Robotics in implant dentistry

Recent advancements in dental robotics have significantly improved precision, efficiency, and clinical outcomes associated with implant placement. Since 2020, the increased adoption of robotic systems such as Yomi, UR robot, THETA, Remebot, ROSA, and Yekebot reflects a growing interest among researchers and clinicians in enhancing surgical accuracy and addressing limitations of conventional methods [40]. Robotic systems have demonstrated superior accuracy in implant placement, with reported deviations in entry point, apex, and angulation consistently lower than those observed in static guide systems as reported in the ITI consensus meta-analysis [10, 40, 41]. This supports the potential of robotics to reduce complications often associated with freehand placement, which relies heavily on the surgeon’s skill and experience [2].

Although static guides and dynamic navigation have contributed to improved accuracy, static guides are limited by their inability to accommodate intraoperative anatomical changes and require high operator expertise [42]. Dynamic navigation systems offer real-time feedback but do not allow for physical manipulation of instruments and come at a higher financial cost [34]. In contrast, robotic-assisted systems integrate mechanical precision with real-time intraoperative guidance, offering enhanced control, visualization, and reproducibility during implant procedures [28, 31, 32].

Yomi was the first robotic device approved by the FDA for dental implant surgery and has been featured in multiple studies as a reliable tool for visual and tactile intraoperative guidance [33, 34, 43]. Other systems such as the UR robot, Remebot, and ROSA have also shown promising in vitro results, particularly in their mechanical stability, clamping precision, and integration with digital workflows [10, 19, 30]. However, the majority of studies to date have been conducted using phantom models or cadaveric specimens [5], which limits generalizability due to the absence of real-world clinical challenges such as patient movement or restricted mouth opening [41].

Notably, robotic systems have been reported to reduce clinician fatigue through stable mechanical arm support [24, 33], enhance intraoperative transparency via real-time navigation data [28], and even support remote collaboration and surgical teaching through network connectivity [22, 44]. While these features contribute to improved procedural quality and patient comfort [22, 44], clinical validation remains essential. The predominance of case reports and in vitro studies in the current body of evidence underscores the urgent need for robust clinical trials to assess long-term outcomes, precision, and safety of these systems in routine practice. Overall, the included studies support the conclusion that robotic-assisted implant surgery offers measurable advantages over traditional techniques in terms of accuracy, control, and workflow efficiency. However, widespread clinical adoption will depend on overcoming current limitations, such as cost, training requirements, and system integration, and will benefit from ongoing developments in artificial intelligence and digitalization [5].

Consideration should also be given to potential errors caused by sudden head movements during implant placement. Studies have reported that muscular fatigue commonly occurs during implant surgeries, leading patients to intermittently close their mouths during the procedure. Although robotic arms incorporate a “follow-up” mechanism, they require time to detect and respond to such movements, which may increase the risk of surgical failure. This highlights the importance of trained operators who can minimize patient head movement throughout the drilling process. Consequently, improving the ability of robotic systems to track rapid movements is essential. In the current review, studies evaluating implant placement in the mandibular region reported this issue, whereas no such errors were noted during maxillary implant placement. This difference may be attributed to the more stable position of the maxilla.

One of the major limitations of robotic systems is their pre-operative preparation, which is time-consuming and contributes to extended surgical duration. Moreover, robotic surgical equipment requires substantial space and a large operating room, as well as additional coordination personnel during the procedure. In addition, the high cost of robotic equipment leads to increased expenses, adding to the financial burden on patients. Furthermore, it may not be suitable for patients to keep their mouths open for prolonged periods throughout the procedure.

Strength and limitations

This scoping review offers a comprehensive overview of the current landscape of robotic applications in both implant and prosthetic dentistry. One of its key strengths is the dual focus on implantology and prosthodontics, allowing for a broader understanding of how robotics is transforming multiple domains of dental care. The review systematically synthesizes findings from a wide range of study designs, including in vitro experiments, preclinical investigations, and early clinical case reports, thereby capturing both foundational technological developments and emerging clinical applications. Another strength lies in the detailed analysis of specific robotic systems such as Yomi, Remebot, ROSA, and UR robot highlighting their capabilities, limitations, and comparative performance. The inclusion of studies across different technological platforms enables clinicians and researchers to assess not only the precision and accuracy of these systems but also their usability, integration into digital workflows, and potential to reduce human error. Additionally, by identifying gaps in the current literature, the review provides guidance for future research directions and clinical validation studies.

However, several limitations must be acknowledged. First, the majority of included studies were preclinical or in vitro, conducted on models or cadaver specimens. These lack the complexity of real-world clinical conditions such as patient movement, anatomical variability, and soft tissue interactions. Second, many of the clinical studies were case reports or small case series, often lacking control groups or long-term follow-up, which restricts conclusions regarding clinical efficacy and patient outcomes. Third, heterogeneity in study designs, outcome measures, robotic platforms, and evaluation metrics posed challenges in synthesizing findings and limited the ability to perform comparative or quantitative analysis. Fourth, there is potential for publication and language bias, as the review included only English-language studies. Finally, the studies in prosthetic dentistry were only published from one country (China) limiting boarder international perspective.

Conclusion

This scoping review highlights the role of robotics in both implant and prosthetic dentistry, demonstrating significant advancements in precision, accuracy, and workflow efficiency. In prosthodontics, emerging robotic applications in tooth preparation and denture tooth arrangement suggest potential for increased standardization and reduced manual error, although these technologies remain in early developmental stages. Robotic-assisted implant surgery, particularly through systems like Yomi, Remebot, and ROSA, has shown promising outcomes in improving surgical accuracy and reducing operator-dependent variability. While the current body of evidence primarily preclinical and descriptive supports the feasibility and benefits of robotic systems, further high-quality clinical research is essential to validate their long-term effectiveness, safety, and cost-efficiency.

Abbreviations

CAIS

Computer-Assisted Implant Surgery

r-CAIS

Robotic computer-assisted implant surgery

CAD/CAM

Computer-aided design and Computer-aided manufacturing

AR

Active robot

SR

Semi-active robot

PR

Passive robot

Authors' contributions

Conceptualization, H.L. and A.J.; methodology, H.L.; validation, H.L. and M.S.; formal analysis, H.L., A.J., and A.A.; Data extraction, H.L. and A.Q.; writing—original draft preparation, A.J. and H.L.; writing—review and editing, S.A.; All authors have read and agreed to the published version of the manuscript.

Funding

No external funding was received in this study.

Data availability

Data will be available at the request of the corresponding author.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.