The construction projects are characterized by intricate relations of cost, schedule, resources, and structural aspects that frequently give rise to uncertainty and project risk. In order to overcome these difficulties, this research suggests a Hybrid Construction Intelligence Framework Network (HCIF-Net) of intelligent construction risk management through combining the BIM data with predictive analytics and hybrid optimization methods. Several heterogeneous datasets, such as records of construction project management, project reports, and IFC-based BIM structural data, are incorporated to form a unified dataset of 1300 project instances. SMOTE balancing is used after the preprocessing and feature engineering to deal with class imbalance in risk categories. The HCIF-Net framework integrates TabNet to learn tables, Bayesian Network to model probabilistic risks dependencies, Temporal Attention LSTM to predict sequential progress, and Monte Carlo simulation to analyse uncertainty. A hybrid MILP-NSGA-II optimization model is employed to reduce the project cost, the time taken to complete, and the total risk to aid decision-making. The results of the experiment show a high level of predictive performance with an accuracy of 98.88% and close-to-perfect ROC-AUC metrics, and BIM-based visualization dashboards offer intuitive insights to monitor and manage construction risk efficiently.
Rui-Xin Li, Reem A. Almenweer, Haytham F. Isleem et al.· Scientific Reports· 0 citations
Vision-Language-Action (VLA) models have become a powerful paradigm for robot manipulation, but training a single generalist policy for heterogeneous robot embodiments remains an open problem. Existing methods have two main limitations. First, they underuse dynamics priors shared across diverse visual and interaction data, limiting cross-embodiment transfer. Second, they require extensive manual preprocessing to convert embodiment-specific actions into a common format. To overcome these limitations, we propose DyPES-VLA, a cross-embodiment VLA that learns shared Dynamics Priors and Embodiment-Specific control. First, we learn shared dynamics priors by training the vision-language model (VLM) with a future-prediction objective on cross-embodiment data, driving the shared query representation to capture object motion, contact, and interaction-induced scene changes. Second, an embodiment-specific Mixture-of-Experts (MoE) action head translates these shared dynamics priors into executable controls directly in each embodiment's native action space, without manually pre-aligning heterogeneous actions into a common format. This head shares attention layers to capture common temporal action structures, while its embodiment-specific feed-forward experts resolve the unique kinematic constraints and control semantics of distinct embodiments. As a generalist policy, our \ourmethod achieves state-of-the-art performance across simulation and real-world evaluations, reaching 98.0% success on LIBERO, 59.25% on RoboCasa-GR1, and 89.02% on RoboTwin~2.0.
Jun-Feng Li, Junjie He, Zhi-De Zhong et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
