Toward the Automation of Search and Rescue Operations: An Algorithm for Finding Missing Lost Persons Living With Dementia Using Drones
Bibliographic record
Abstract
Unmanned Aerial Vehicles (UAV) are now used in many applications. The focus in this research work is on their use for public safety, specifically in Search and Rescue (SAR) operations involving Lost Persons Living With Dementia (LPLWD). When it comes to saving lives, there are many human factors associated with UAV operations that impact the performance of expert human SAR teams that could be improved through some form of automation. These include familiarity with the search location, and tasks associated with piloting and search/flight management during SAR operations. An LPLWD may not be interested in assisting in their own rescue as they may not know they are lost. As such, it has been observed that they tend to keep walking until they are faced with an obstacle that bars their continued progress. The approach presented in this research work focuses on developing a people-finding method to identify high-probability locations where an LPLWD might be found through informed, behavior-based analysis of the search location, and then developing an algorithm to fly a UAV to the vicinity of these high-probability locations. The algorithm was validated through field testing. The results from both the data collection process and the field tests indicated that there were efficiencies in using the drone, which enhanced the probability of finding the lost person alive. An informed process, involving automated approaches using R software to scrub and augment the data by adding any missing values in the dataset, helped to clarify the behaviour of the lost person and determine what significant variables enhanced their survivability. Linear regression was utilized to acquire the correlations between the numeric values in the database. Logistic regression and classification models were used to investigate survivability. The final goal of this work is to address the possibility of automating certain types of SAR. The outcome will also inform the design of use cases that could be used in various test environments.
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 | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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".