End-to-End Prediction of Parcel Delivery Time with Deep Learning for\n Smart-City Applications
Bibliographic record
Abstract
The acquisition of massive data on parcel delivery motivates postal operators\nto foster the development of predictive systems to improve customer service.\nPredicting delivery times successive to being shipped out of the final depot,\nreferred to as last-mile prediction, deals with complicating factors such as\ntraffic, drivers' behaviors, and weather. This work studies the use of deep\nlearning for solving a real-world case of last-mile parcel delivery time\nprediction. We present our solution under the IoT paradigm and discuss its\nfeasibility on a cloud-based architecture as a smart city application. We focus\non a large-scale parcel dataset provided by Canada Post, covering the Greater\nToronto Area (GTA). We utilize an origin-destination (OD) formulation, in which\nroutes are not available, but only the start and end delivery points. We\ninvestigate three categories of convolutional-based neural networks and assess\ntheir performances on the task. We further demonstrate how our modeling\noutperforms several baselines, from classical machine learning models to\nreferenced OD solutions. Specifically, we show that a ResNet architecture with\n8 residual blocks displays the best trade-off between performance and\ncomplexity. We perform a thorough error analysis across the data and visualize\nthe deep features learned to better understand the model behavior, making\ninteresting remarks on data predictability. Our work provides an end-to-end\nneural pipeline that leverages parcel OD data as well as weather to accurately\npredict delivery durations. We believe that our system has the potential not\nonly to improve user experience by better modeling their anticipation but also\nto aid last-mile postal logistics as a whole.\n
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 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 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 | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 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".