Data issues in industrial AI systems: A meta-review and research strategy
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
In the era of Industry 4.0, artificial intelligence (AI) is assumed to play an increasingly pivotal role within industrial systems. Despite the recent trend within various industries to adopt AI, the actual adoption of AI is not as developed as perceived. A significant factor contributing to this lag is the data issues in AI implementation. How to address these data issues stands as a significant concern confronting both industry and academia. Thus, this study conducts a comprehensive meta-review of data issues and corresponding methods in industrial AI. Eighty-two data issues are identified and categorized into seven stages of the data lifecycle. To supplement the existing research that focuses more on data issues arising in historical data, this study subsequently discusses the management of real-time sensor data and expert domain knowledge. Meanwhile, it proposes a model-aware data preparation approach, which integrates the data characteristics with specific AI model requirements to enhance data usability and algorithm alignment. This approach is further integrated into a conceptual framework that combines managerial and technical perspectives for systematically resolving data issues. The framework provides actionable insights and a systematic method for AI practitioners and industrial system developers to anticipate and address data-related challenges. Finally, the study highlights future research directions. This study advances the existing body of knowledge, supports a seamless transition from traditional model-centric AI to data-centric AI, and offers practical guidelines for professionals navigating the complexities of achieving data excellence in industrial AI applications. • In AI projects, it is critical to consider industrial needs and data usability. • The successful implementation of AI is hindered by various data issues. • Data lifecycle theory offers a structured framework for examining data issues. • Aligning data nature with specific AI model requirements is critical.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it