Office building occupancy monitoring through image recognition sensors
Bibliographic record
Abstract
In the Architecture, Engineering, construction and Operations (AEcO) there is a growing interest in the use of the building Information modelling (bIm). Through integration of information and processes in a digital model, bIm can optimise resources along the lifecycle of a physical asset. Despite the potential savings are much higher in the operational phase, bIm is nowadays mostly used in design and construction stages and there are still many barriers hindering its implementation in Facility management (Fm). Its scarce integration with live data, i.e. data that changes at high frequency, can be considered one of its major limitations in Fm. The aim of this research is to overcome this limit and prove that buildings or infrastructures operations can benefit from a digital model updated with live data. The scope of the research concerns the optimisation of Fm operations. The optimisation of operations can be further enhanced by the use of maintenance smart contracts allowing a better integration between users' behaviour and maintenance implementation. In this case study research, the Image recognition (Imr), a type of Artificial Intelligence (AI), has been used to detect users' movements in an office building, providing real time occupancy data. This data has been stored in a bIm model, employed as single reliable source of information for Fm. This integration can enhance maintenance management contracts if the bIm model is coupled with a smart contract. Far from being a comprehensive case study, this research demonstrates how the transition from bIm to the Asset Information model (AIm) and, finally, to the Digital Twin (i.e. a near-real-time digital clone of a physical asset, of its conditions and processes) is desirable because of the outstanding benefits that have already been measured in other industrial sectors by applying the principles of Industry 4.0.
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 | 0.000 |
| 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 | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
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".