Slope failures in open-pit mining pose significant operational and safety issues, underscoring the importance of implementing effective stability monitoring frameworks for early hazard detection to allow for timely intervention and risk mitigation. This systematic review presents a comprehensive synthesis of existing and emerging methods and technologies used for slope stability monitoring in open-pit mining, including both remote sensing and in situ methods, as well as advanced technologies, such as Artificial Intelligence (AI), the Internet of Things (IoT), and Wireless Sensor Networks (WSNs). Using the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) 2020 guidelines, a total of 49 studies were selected from a collection of four engineering databases, and a comparative analysis was conducted to determine the underlying differences between the various methods for open-pit slope stability monitoring in terms of their performance across key attributes, such as monitoring accuracy, spatial and temporal coverage, operational complexity, and economic viability. Their juxtaposition highlighted the notion that no universally optimal slope stability monitoring system exists, due to a series of compromises that arise as a result of inherent technological limitations and site-specific constraints. Notably, remote sensing methods offer large-scale, non-intrusive monitoring, but are often limited by environmental factors and data acquisition infrequency, whereas in situ methods provide high precision, but suffer from limited spatial coverage and scalability. This review further highlights the capacity of emerging methods and technologies to address these limitations, providing suggestions for future research directions involving the integration of multiple sensing technologies for the enhancement of monitoring capabilities. This study provides a consolidated knowledge base on open-pit slope stability monitoring methods, technologies, and techniques, to guide the development of integrated, cost-effective, and scalable slope monitoring solutions that enhance mine safety and efficiency.

Open-pit mining operations rely heavily on visual inspection to identify indicators of slope instability such as surface cracks. Early identification of these geotechnical hazards enables timely safety interventions to protect both workers and assets in the event of slope failures or landslides. While computer vision (CV) approaches offer a promising avenue for autonomous crack detection, their effectiveness remains constrained by the scarcity of labelled geotechnical datasets. Deep learning (DL)-based models, in particular, require large amounts of representative training data to generalize to unseen conditions; however, collecting such data from operational mine sites is limited by safety, cost, and data confidentiality constraints. To address this challenge, this study proposes a novel hybrid game engine–generative artificial intelligence (AI) framework for large-scale dataset generation without requiring real-world training data. Leveraging a parameterized virtual environment developed in Unreal Engine 5 (UE5), the framework generates realistic images of open-pit surface cracks and enhances their fidelity and diversity using StyleGAN2-ADA. The synthesized datasets were used to train the YOLOv11 real-time object detection model and evaluated on a held-out real-world dataset of open-pit slope imagery to assess the effectiveness of the proposed framework in improving model generalizability under extreme data scarcity. Experimental results demonstrated that models trained using the proposed framework consistently outperformed the UE5 baseline, with average precision (AP) at intersection over union (IoU) thresholds of 0.5 and [0.5:0.95] increasing from 0.792 to 0.922 (+16.4%) and 0.536 to 0.722 (+34.7%), respectively, across the best-performing configurations. These findings demonstrate the effectiveness of hybrid generative AI frameworks in mitigating data scarcity in CV applications and supporting the development of scalable automated slope monitoring systems for improved worker safety and operational efficiency in open-pit mining.

Pixel-level annotation remains a major bottleneck for semantic segmentation, motivating methods that synthesize image–label pairs directly from generative models. Prior synthetic dataset generators typically obtain pseudo-labels from cross-attention maps or learned decoders over generative features; however, recent text-to-image (T2I) models increasingly use multimodal diffusion transformers (MM-DiTs), where concept localization is no longer exposed through a single cross-attention pathway but instead distributed across many layers and attention heads. Existing MM-DiT localization methods address this by aggregating saliency across heads, but we observe that this averaging can dilute clean target localizers due to attention head heterogeneity. We introduce HEADHUNTER, a training-free segmentation framework that uses aggregate concept saliency as a self-guided proxy to select the single attention head that best localizes a queried textual concept, yielding cleaner segmentation masks. We then use HEADHUNTER to turn target classes into training data automatically: a large language model (LLM) diversifies prompts, an MM-DiT generates images, HEADHUNTER produces pseudo-labels, and a vision language model (VLM) verifies each image–mask pair before acceptance. HEADHUNTER achieves strong zero-shot segmentation performance (81.2 mIoU on PASCAL VOC2012 and 73.2 mIoU on ImageNet-Segmentation), outperforming head aggregation and other interpretability methods. Using only our generated image–label pairs, we train segmentation models which reach 64.9 mIoU on VOC2012 validation, matching or outperforming comparable synthetic dataset generators and showing that the proposed pipeline produces effective dense labels without human intervention.

Pixel-level annotation remains a major bottleneck for semantic segmentation, motivating methods that synthesize image-label pairs directly from generative models. Prior synthetic dataset generators typically obtain pseudo-labels from cross-attention maps or learned decoders over generative features; however, recent text-to-image (T2I) models increasingly use multimodal diffusion transformers (MM-DiTs), where concept localization is no longer exposed through a single cross-attention pathway but instead distributed across many layers and attention heads. Existing MM-DiT localization methods address this by aggregating saliency across heads, but we observe that this averaging can dilute clean target localizers due to attention head heterogeneity. We introduce HEADHUNTER, a training-free, open-vocabulary segmentation framework that uses aggregate concept saliency as a self-guided proxy to select the single attention head that best localizes a queried textual concept, yielding cleaner segmentation masks. We then use HEADHUNTER to turn target classes into training data automatically: a large language model (LLM) diversifies prompts, an MM-DiT generates images, HEADHUNTER produces pseudo-labels, and a vision language model (VLM) verifies each image-mask pair before acceptance. HEADHUNTER achieves strong zero-shot segmentation performance (81.2 mIoU on PASCAL VOC2012 and 73.2 mIoU on ImageNet-Segmentation), outperforming head aggregation and other interpretability methods. Using only our generated image-label pairs, we train segmentation models which reach 64.9 mIoU on VOC2012 validation, matching or outperforming comparable synthetic dataset generators and showing that the proposed pipeline produces effective dense labels without human intervention.

Open-pit mining relies heavily on visual inspection to identify indicators of slope instability such as surface cracks. Early identification of these geotechnical hazards allows for the implementation of safety interventions to protect both workers and assets in the event of slope failures or landslides. While computer vision (CV) approaches offer a promising avenue for autonomous crack detection, their effectiveness remains constrained by the scarcity of labelled geotechnical datasets. Deep learning (DL) models require large amounts of representative training data to generalize to unseen conditions; however, collecting such data from operational mine sites is limited by safety, cost, and data confidentiality constraints. To address this challenge, we propose a hybrid game engine—generative artificial intelligence (AI) framework for large-scale dataset synthesis. Leveraging a parameterized virtual environment developed in Unreal Engine 5 (UE5), the framework captures realistic images of open-pit surface cracks and enriches their visual diversity using StyleGAN2-ADA. The resulting datasets were used to train the YOLOv11 real-time object detection model and evaluated on a real-world dataset of open-pit slope imagery to assess the effectiveness of the proposed framework in improving CV model generalizability under extreme data scarcity. Experimental results demonstrated that models trained on the proposed framework substantially outperformed the UE5 baseline, with average precision (AP) at intersection over union (IoU) thresholds of 0.5 and [0.5:0.95] increasing from 0.403 to 0.922 and 0.223 to 0.722 respectively, accompanied by a reduction in missed detections from 95 to eight for the best-performing configurations. These findings demonstrate the potential of hybrid generative AI frameworks to mitigate data scarcity in CV applications and support the development of scalable automated slope monitoring systems for improved worker safety and operational efficiency in open-pit mining.

Automatic crack segmentation in open-pit mining is constrained by the scarcity of representative, labeled datasets. Vision foundation models (VFMs) offer transferable visual representations, but their training-free performance on mining imagery and the benefits of synthetic-only adaptation remain insufficiently established. This study evaluates training-free VFM pipelines and investigates whether synthetic supervision improves their transfer to real-world crack imagery. SynthMineCrack is introduced as a diverse open-pit mining crack dataset, supporting model training and hyperparameter selection without real imagery or manual annotations. SAM 3 is adapted using low-rank adaptation (LoRA), layer normalization (LN) tuning, and adapters, individually and jointly, and compared with task-specific models trained on the same synthetic data. Evaluation on both in- and out-of-domain datasets shows that SAM 3 achieves the strongest training-free performance, while joint adaptation of all three operators improves intersection-over-union (IoU) on the mining evaluation dataset by 18.25 percentage points (pp). This configuration achieves a mean IoU of 69.39% across the five datasets, exceeding SegFormer, the strongest task-specific model, by 10.77 pp. Diagnostic analyses associate these gains with more complete crack recovery while the pretrained visual representation remains largely preserved. This configuration updates only 0.326% of the instantiated model’s parameter count and incurs little additional inference overhead. Together, these findings support synthetic-only adaptation as a practical means of improving SAM 3 for real-world crack segmentation while retaining inference requirements close to those of the base model.