Trajectory Prediction of Mobile Construction Resources Toward Pro-active Struck-by Hazard Detection
Bibliographic record
Abstract
Trajectory Prediction of Mobile Construction Resources Toward Pro-active Struck-by Hazard Detection Daeho Kim, Meiyin Liu, Sanghyun Lee and Vineet R. Kamat Pages 982-988 (2019 Proceedings of the 36th ISARC, Banff, Canada, ISBN 978-952-69524-0-6, ISSN 2413-5844) Abstract: In construction, unanticipated struck-by hazards often arise, which have resulted in a significant number of construction fatalities. To address this problem, many studies have attempted to automate proximity monitoring and struck-by hazard detection using various technologies, such as wireless sensors and computer vision methods. While this technology focuses on understanding what is happening as hazards arise, it is not equipped to detect future hazards. In impending situations, detecting current hazards may not provide enough time for workers to take evasive actions. To address this challenge this study develops a trajectory prediction model for mobile construction resources. Specifically, this study conducts hyper-parameter tuning of a deep neural network, called Social Generative Adversarial Network to develop a prediction model capable of predicting more than five seconds. Further, a test on a real construction operations data follows to validate developed models' trajectory prediction accuracy. As a result, a developed model could achieve promising accuracy: the average displacement error and the final displacement error were 0.78 and 1.27 meters, respectively. The trajectory prediction allows for detecting future hazards, which will support pro-active intervention in hazardous situations. It will ultimately contribute to promoting a safer working environment for construction workers. Keywords: Struck-by hazard; Pro-active intervention; Trajectory prediction; Deep neural network; Hyper-parameter tuning DOI: https://doi.org/10.22260/ISARC2019/0131 Download fulltext Download BibTex Download Endnote (RIS) TeX Import to Mendeley
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".