A systematic review of immersive virtual reality for industrial skills training
Open accessThe DOI matches its Crossref record; 123 of the 123 references it cites were found in a registry or an index; 31 of the 122 sources its claims cite were checked for support.
Abstract
Virtual reality (VR) training offers the capability to industrial workers to acquire skills and address complex tasks by immersing them in a safe and controlled virtual environment. Immersive VR (IVR) training is adopted in many diverse settings, yet little systematic work currently exists on how researchers have applied it for industrial skills training and if it holds the potential to be applied remotely. In this review, 78 representative studies were analysed to answer three key questions: Is IVR an effective training method for industrial skills training? How is research in this field applied? And how can we make IVR training more effective and applicable for remote training? We can testify that IVR is a promising training method with high effectiveness scores. However, our analysis has uncovered several gaps in the application of IVR training, like the lack of learning theories in the design process and limited metrics beyond time and scores. Additionally, our review also exposed unexplored but intriguing avenues of research, like the utilisation of biosensors for users’ data collection, haptics that increases realism and applications with remote training potential.
Overview
CoChat Summary
TL;DR
This PRISMA systematic review analyses 78 representative studies on immersive virtual reality (IVR) for industrial skills training from 2010–2020. It finds IVR promising and often effective, but identifies gaps in learning-theory grounding, measurement beyond time and scores, and remote-training applicability.
Key findings
- IVR was reported as effective in 29 of the 40 publications that compared it with another modality; 6 found it not as effective, 5 were inconclusive, and none indicated IVR was unsuitable for training industrial workers.
- Procedural skills were the most trained skill type at 45% (N=35), followed by perceptual-motor skills at 33% (N=26), decision-making at 17% (N=13), and spatial skills at 5% (N=4 in the results section; the discussion says N=5).
- Only 14% (N=11) of publications explicitly mentioned a learning theory, with 13 learning theories identified; common measures were task completion time (42%, N=33), scoring metrics (N=41), cybersickness (18%, N=14), usability (20%, N=16), task load (17%, N=13), and immersion (18%, N=14).
- Most applications were classified as autonomous training (N=60), while remote guidance (N=3) and remote peer collaboration (N=1) were rare; HTC VIVE was the most used HMD (45%, N=35), followed by Oculus Rift (26%, N=20), and 35% (N=27) used some form of haptics.
Methods
- Systematic review following PRISMA, searching Scopus, Web-of-Science, and IEEE-Xplore for peer-reviewed papers from January 2010 to December 2020; the initial 13,715 publications were filtered to 78 representative studies.
- Filtering included duplicate removal, Article Citation Rate ranking with a cut-off of 2, semi-automatic exclusion based on ability, technology, domain/setting, quality, and other terms, then manual voting by three authors with a Fleiss Kappa of 0.578.
- A coding framework addressed six research questions covering skills, learning theories, research designs, data collection/analysis, HMDs/biosensors/haptics, effectiveness, and remote-training applicability.
Limitations
- The paper states that only peer-reviewed publications were included, so industry reports, books, and cases may have been excluded; the GICS-inspired categorisation is necessary but restrictive; and the IVR user was defined as a healthy neurotypical adult without profound mental/physical disabilities or illness, limiting inclusivity and literature count.
- Evident in the text: several reported counts are inconsistent across sections (e.g., spatial skills N=4 vs. N=5; biosensors 18% N=14 vs. 14% N=11; haptics 35% N=27 vs. 32% N=25; effectiveness N=29 vs. “one-third”), and the review is descriptive rather than meta-analytic, with no clear recipe for what makes IVR effective.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
4 Valid · 24 Not valid · 55 Not checked · 147 Own results and interpretations
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“- Constructivist Learning Theory: Akanmu et al. use Constructivist learning theory to design their IVRbased posture training system based on the learner's current and previous experience [65].”
What the sources say
- Could not verifycites [65] sourceExperimental — accuracy not yet measured
Our copy’s Theoretical implications (p. 13) discuss cognitive theory of multimedia learning and pre-training, while Empirical contributions (p. 11) describe a VR learning experience set inside a cell. Searches for Akanmu, posture training, and constructivist learning theory found no passage addressing the attributed system or its design. (LLM, conf=0.75)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“whereas Škola, Tinková [55] created an EEG-based adaptive training system for increasing the user affect in a gamified virtual environment to provide motor imagery-related biofeedback.”
What the sources say
- Could not verifycites [55] sourceExperimental — accuracy not yet measured
The paper describes Liang, Zhou, and Gao’s virtual-reality game for underground mine safety training (p. 1) and performance-based adaptation of its training content (p. 2). Searches for Škola and Tinková, EEG, motor imagery, and biofeedback found no passage addressing the system attributed to them. (LLM, conf=0.77)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“The other brain-signal sensing technology in the database is in Shi et al. [84], where a functional near-infrared spectroscopy (fNIRS) device (a NIRSportTM worn on the head) is used along with an HMD integrated eye tracker (Tobii) to measure stress levels among users who were being trained on industrial shutdown maintenance.”
What the sources say
- Could not verifycites [84] sourceExperimental — accuracy not yet measured
The Recording of Physiological Signals section (p. 5) describes Putze et al.’s use of a NeXus 10 device with respiration, blood-volume-pressure, and skin-conductance sensors. Searches for Shi, NIRSport, fNIRS, Tobii, and industrial shutdown maintenance found no passage about the study described in the claim. The paper therefore does not verify what Shi et al. used. (LLM, conf=0.76)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“certain design choices for increasing IVR realism may improve training success in IVR”
What the sources say
- Not matchedcites [42], [102], [112], [113], [87]
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Supportedcites [55] source
The Conclusion (p. 9) says realistic, immediate physics feedback makes the training experience more memorable and that immersive design helps trainees become familiar with a dangerous environment. It also reports better performance for VR trainees than for video trainees and cautiously suggests immersion as a reason. This supports the claim's qualified link between realism-related design choices and training success. (LLM, conf=0.88)
- Not matchedcites [81], [64]
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Not matched
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Only one publication indicated that gamified features in their IVR training led to better engagement [67].”
What the sources say
- Could not verifycites [67] sourceExperimental — accuracy not yet measured
The PRISMA Statement section describes a checklist and flow diagram for reporting systematic reviews, not findings about IVR training. Searches for gamified IVR training and engagement found no passage addressing the claim; the paper’s copy does not establish that one publication reported better engagement. (LLM, conf=0.78)
hedgedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
88 more claims · 1 with their cited sources checked
Create a free account to see all 230 claims and every source verdict.
- Probably not a substantial claim. 66
- It may state two separate claims as one. 85
- The quote only partly states this claim. 2
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ 5.4.2 Use of biosensors
“18% (N=14) of the publications used biosensors measuring brain signals (N=8), heart rate (N=3), eye tracking (N=6), temperature (N=1) and skin conductance (N=1).”
§ How can we improve IVR training for the industry and make it applicable for remote training?
“However, this opportunity to use biosensors was not extensively evident in our data set, as only 14% (N=11) of the publications used these data-gathering devices.”
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ 5.4.3 Haptic feedback
“35% of the publications (N=27) used a variation of haptic feedback.”
§ How can we improve IVR training for the industry and make it applicable for remote training?
“32% of the publications (N=25) used a variant of haptic feedback.”
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
“In our case, the papers that reported a learning theory (N=14), 13 examples were identified, constituting a comprehensive, valuable list for IVR designers and developers (depicted in section 5.2).”
Reported 14 · recomputed 17.948718
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ 5.4.2 Use of biosensors
“18% (N=14) of the publications used biosensors”
§ How can we improve IVR training for the industry and make it applicable for remote training?
“only 14% (N=11) of the publications used these data-gathering devices.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ 5.2 Learning Theories ( RQ2: Which learning theories are utilised in IVR industrial skills training? )
“In 14% (N=11) of the total number of publications, the authors explicitly mentioned 13 learning theories”
§ How do we conduct research in the IVR for the industrial training field?
“In our case, the papers that reported a learning theory (N=14), 13 examples were identified, constituting a comprehensive, valuable list for IVR designers and developers (depicted in section 5.2).”
13 more findings
Create a free account to see every finding of this check.
Citations
Cites
Preferred Reporting Items for Systematic Reviews and Meta-Analyses: the PRISMA statement
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & for the PRISMA Group (2009) · BMJ · doi:10.1136/bmj.b2535 · Cited by 83,823
Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & The PRISMA Group (2009) · PLoS Medicine · doi:10.1371/journal.pmed.1000097 · Cited by 64,042 · cited 2×
“The fi ltering process was carried out following the Pre-ferred Reporting Items for Systematic Literature Reviews and Meta-Analyses (PRISMA) framework (Moher et al. 2009).”
MethodDevelopment of NASA-TLX (Task Load Index): Results of Empirical and Theoretical Research
(1988) · Advances in Psychology · doi:10.1016/s0166-4115(08)62386-9 · Cited by 15,003 · cited 1×
“For the task load, the most popular choices are the NASA-TLX (task load index) (Hart and Staveland 1988), the instantaneous self-assessment (ISA) technique (Tattersall and Foord 1996) and the SIM-TLX (simulation task load index) (Harris, Wilson, and Vine 2019).”
MethodMeasuring nominal scale agreement among many raters.
Fleiss, J. L. (1971) · Psychological Bulletin · doi:10.1037/h0031619 · Cited by 8,733 · cited 1×
“A Fleiss Kappa (Fleiss 1971) inter-rater reliability score of 0.578 was calculated on the 802 ratings by three judges, showing moderate agreement (0.41 – 0.60).”
MethodInfluentialMeasuring Presence in Virtual Environments: A Presence Questionnaire
Witmer, B. G., & Singer, M. J. (1998) · Presence: Teleoperators and Virtual Environments · doi:10.1162/105474698565686 · Cited by 6,371 · cited 1×
“Finally, immersion metrics include the presence questionnaire (Witmer and Singer 1998) and the Igroup presence questionnaire (IPQ) (Schubert, Friedmann, and Regenbrecht 2001).”
MethodSimulator Sickness Questionnaire: An Enhanced Method for Quantifying Simulator Sickness
Kennedy, R. S., Lane, N. E., Berbaum, K. S., & Lilienthal, M. G. (1993) · The International Journal of Aviation Psychology · doi:10.1207/s15327108ijap0303_3 · Cited by 5,457 · cited 1×
“Unspeci fi ed/generic The industry is not mentioned, abstracted tasks which hold the potential to be applied to industrial use cases. cybersickness are the simulator sickness questionnaire (SSQ) (Kennedy et al. 1993) and the motion sickness scale (Keshavarz and Hecht 2011).”
BackgroundRubber hands ‘feel’ touch that eyes see
Botvinick, M., & Cohen, J. (1998) · Nature · doi:10.1038/35784 · Cited by 4,410
Radianti, J., Majchrzak, T. A., Fromm, J., & Wohlgenannt, I. (2020) · Computers & Education · doi:10.1016/j.compedu.2019.103778 · Cited by 3,585 · cited 9×
“As of late, a new innovative wave of interactive immersive virtual reality (IVR) systems is being utilised for industrial training (Radianti et al. 2020).”
BackgroundMethodResultInfluentialJournal on Educational Psychology
Winston, J., Mandernach, J., Winsor, D. L., Sinnott, J. D., & Gujjar, A. A. (2014) · Cited by 2,947
Memory: A Contribution to Experimental Psychology
Ebbinghaus, H. (2013) · Annals of Neurosciences · doi:10.5214/ans.0972.7531.200408 · Cited by 2,827 · cited 1×
“For example, the Ebbinghaus model (Ebbinghaus 2013) suggests that students forget around 75% of what they learned within a few days in non-VR studies.”
Method
Research works printed but not linked: 6
- Biggs S. J., 2002, Handbook of Virtual Environments, 93
- Eckert Daniel, 2020, The Effectiveness of Virtual Reality Soft Skills Training in the Enterprise
- Mayer R. E., 1998, Journal of Educational Psychology, 91, 358
- Mendes F. A. D. S., 2012, Physiotherapy (United Kingdom), 98, 217
- Morrison T., 2011, Boiling Point? The Skills Gap in US Manufacturing
- S&P Global and MSCI, 2018, The Global Industry Classification Standard (GICS®)
Cited by
Alpala, L. O., Quiroga-Parra, D. J., Torres, J. C., & Peluffo-Ordóñez, D. H. (2022) · Applied Sciences · doi:10.3390/app12126258 · Cited by 164
“It is designed to simulate real-life, and its two essential characteristics are immersion and presence [26].”
BackgroundEvaluating the effectiveness of virtual reality for safety-relevant training: a systematic review
Stefan, H., Mortimer, M., & Horan, B. (2023) · Virtual Reality · doi:10.1007/s10055-023-00843-7 · Cited by 195
“Radhakrishnan et al. (2021) performed a systematic review of immersive VR for skills training, which also included a wide range of industries.”
BackgroundReview of smart water management: IoT and AI in water and wastewater treatment
Dada, M. A., Majemite, M. T., Obaigbena, A., Daraojimba, O. H., Oliha, J. S., & Nwokediegwu, Z. Q. S. (2024) · World Journal of Advanced Research and Reviews · doi:10.30574/wjarr.2024.21.1.0171 · Cited by 112
“This technology enables more accurate modeling, predictive analytics, and scenario testing, supporting dynamic decision-making for optimizing water treatment processes and system performance (Hou et al., 2017; Radhakrishnan, Koumaditis, & Chinello, 2021).”
MethodImpact of Latency on QoE, Performance, and Collaboration in Interactive Multi-User Virtual Reality
Damme, S. V., Sameri, J., Schwarzmann, S., Wei, Q., Trivisonno, R., Turck, F. D., & Vega, M. T. (2024) · Applied Sciences · doi:10.3390/app14062290 · Cited by 59
Establishment of Virtual-Reality-Based Safety Education and Training System for Safety Engagement
Seo, H. J., Park, G. M., Son, M., & Hong, A. J. (2021) · Education Sciences · doi:10.3390/educsci11120786 · Cited by 51
“[24] Immersion Theory and Embodied learning Industrial skill • 3D environment • Educational factors: Industry classifications and learning of processes and sequences”
BackgroundInfluential
Metadata
- Authors
- Radhakrishnan, U., Koumaditis, K., & Chinello, F.
- Affiliations
- Aarhus University (DK)
- Corresponding author
- Radhakrishnan, Unnikrishnan
- Venue
- Behaviour & Information Technology
- Publisher
- Informa UK Limited
- Published
- 2021-07-31
- Volume
- 40
- Issue
- 12
- Pages
- 1310-1339
- ISSN
- 0144-929X (ISSN-L)
- 1362-3001
- OpenAlex
- W3205141098
- Cited by
- 200
- References
- 122
- Citations by year
- 2026: 46 · 2025: 64 · 2024: 41 · 2023: 35 · 2022: 12 · 2021: 2
- FWCI
- 2.19
- Type
- journal article
- Language
- en
- Open access
- green
- pure.au.dk
- Licence
- cc-by-nc
- Locations
- Behaviour and Information Technology (publishedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- Topics
- Physical Sciences › Computer Science › Human-Computer Interaction › Virtual Reality Applications and Impacts
- Physical Sciences › Engineering › Control and Systems Engineering › Simulation and Modeling Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Augmented Reality Applications
- Keywords
- immersive virtual reality, remote training, training program design, learning theory, training metrics, biosensors, haptics, industrial workers
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-09-28, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-09-28 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: A systematic review of immersive virtual reality for industrial skills training. https://app.cochat.ai/papers/10.1080/0144929x.2021.1954693
- The paper
- Radhakrishnan, U., Koumaditis, K., & Chinello, F. (2021). A systematic review of immersive virtual reality for industrial skills training. Behaviour & Information Technology. https://doi.org/10.1080/0144929x.2021.1954693
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.