Artificial intelligence in the construction industry: A review of present status, opportunities and future challenges

Journal of Building Engineering, Journal Year: 2021, Volume and Issue: 44, P. 103299 - 103299

Published: Oct. 6, 2021

The growth of the construction industry is severely limited by myriad complex challenges it faces such as cost and time overruns, health safety, productivity labour shortages. Also, one least digitized industries in world, which has made difficult for to tackle problems currently faces. An advanced digital technology, Artificial Intelligence (AI), revolutionising manufacturing, retail, telecommunications. subfields AI machine learning, knowledge-based systems, computer vision, robotics optimisation have successfully been applied other achieve increased profitability, efficiency, safety security. While acknowledging benefits applications, numerous are relevant still exist industry. This study aims unravel examine techniques being used identify opportunites applications A critical review available literature on activity monitoring, risk management, resource waste was conducted. Furthermore, opportunities were identified presented this study. provides insights into key applies construction-specific challenges, well pathway realise acrueable

Language: Английский

---

Deep learning in the construction industry: A review of present status and future innovations

Journal of Building Engineering, Journal Year: 2020, Volume and Issue: 32, P. 101827 - 101827

Published: Sept. 19, 2020

The construction industry is known to be overwhelmed with resource planning, risk management and logistic challenges which often result in design defects, project delivery delays, cost overruns contractual disputes. These have instigated research the application of advanced machine learning algorithms such as deep help diagnostic prescriptive analysis causes preventive measures. However, publicity created by tech firms like Google, Facebook Amazon about Artificial Intelligence applications unstructured data not end field. There abound many learning, particularly within sector areas site planning management, health safety prediction, are yet explored. overall aim this article was review existing studies that applied prevalent structural monitoring, safety, building occupancy modelling energy demand prediction. To best our knowledge, there currently no extensive survey techniques industry. This would inspire future into how apply image processing, computer vision, natural language processing numerous Limitations black box challenge, ethics GDPR, cybersecurity cost, can expected researchers practitioners when adopting some these were also discussed.

Language: Английский

---

Convolutional neural network-based data anomaly detection method using multiple information for structural health monitoring

Structural Control and Health Monitoring, Journal Year: 2018, Volume and Issue: 26(1), P. e2296 - e2296

Published: Nov. 29, 2018

Structural health monitoring (SHM) is used worldwide for managing and maintaining civil infrastructures. SHM systems have produced huge amounts of data, but the effective monitoring, mining, utilization this data still need in-depth study. generally includes multiple types anomalies caused by sensor faults or system malfunctions that can disturb structural analysis assessment. In routine pre-processing, signal processing techniques are required to detect anomalies, respectively, which inefficient. The large variations extracted features from massive make anomaly detection prone be over-processed under-processed. Even with expert intervention, parameter tuning, associated preprocessing methods, a challenge, making procedure expensive addition, one technique frequently mis-detects other anomaly. work, we focus on in stage pre-processing little work has been done based real-world continuous multiclass anomalies. We proposed novel method convolutional neural network (CNN) imitates human vision decision making. First, split raw time series into sections, visualized frequency domain, respectively. Then each section's images were stacked as single dual-channel image labeled according graphical (multi-2D space expression). Second, CNN was designed trained classification, during descriptions representations anomalies' learned convolution. To validate our considered effects balanced imbalanced training sets ratios actual acceleration an long-span cable-stayed bridge. results show could multipattern efficiently high accuracy. dual-information CNN-based design makes process readily scalable, faster, more accurate, thereby providing perspective strong potential preprocessing.

Language: Английский

---

Computer Vision Techniques in Construction: A Critical Review

Archives of Computational Methods in Engineering, Journal Year: 2020, Volume and Issue: 28(5), P. 3383 - 3397

Published: Oct. 19, 2020

Language: Английский

---

Machine learning in construction: From shallow to deep learning

Developments in the Built Environment, Journal Year: 2021, Volume and Issue: 6, P. 100045 - 100045

Published: Feb. 21, 2021

The development of artificial intelligence technology is currently bringing about new opportunities in construction. Machine learning a major area interest within the field intelligence, playing pivotal role process making construction "smart". application machine has potential to open up an array such as site supervision, automatic detection, and intelligent maintenance. However, implementation faces range challenges due difficulties acquiring labeled data, especially when applied highly complex environment. This paper reviews history from shallow deep its applications strengths weaknesses have been analyzed order foresee future direction this sphere. Furthermore, presents suggestions which may benefit researchers terms combining specific knowledge domains with algorithms so develop dedicated network models for industry.

Language: Английский

---

Automated excavators activity recognition and productivity analysis from construction site surveillance videos

Automation in Construction, Journal Year: 2019, Volume and Issue: 110, P. 103045 - 103045

Published: Dec. 10, 2019

Language: Английский

---

End-to-end vision-based detection, tracking and activity analysis of earthmoving equipment filmed at ground level

Automation in Construction, Journal Year: 2019, Volume and Issue: 105, P. 102811 - 102811

Published: May 23, 2019

Language: Английский

---

Dynamic prediction for attitude and position in shield tunneling: A deep learning method

Automation in Construction, Journal Year: 2019, Volume and Issue: 105, P. 102840 - 102840

Published: May 28, 2019

Language: Английский

---

Dataset and benchmark for detecting moving objects in construction sites

Automation in Construction, Journal Year: 2020, Volume and Issue: 122, P. 103482 - 103482

Published: Nov. 30, 2020

Language: Английский

---

Computer vision technologies for safety science and management in construction: A critical review and future research directions

Safety Science, Journal Year: 2021, Volume and Issue: 135, P. 105130 - 105130

Published: Jan. 9, 2021

Language: Английский

---