Predicting the Last Mile: Route-Free Prediction of Parcel Delivery Time with Deep Learning for Smart-City Applications
Bibliographic record
Abstract
The massive acquisition of parcel data motivates postal operators to foster the development of predictive systems for better customer service. Predicting delivery after the parcels are sent out of the final depot, referred to as \textit{last-mile} prediction, deals with complicating factors such as traffic, drivers' behaviors, and weather conditions. Our work provides an end-to-end neural pipeline that leverages parcel and weather data to accurately predict delivery durations. We present our solution under the IoT paradigm and discuss its feasibility on a cloud-based architecture as a smart city application. We utilize a route-free origin-destination (OD) formulation, only relying on the delivery start and end points. We use a large-scale real-world dataset provided by Canada Post, containing last-mile information of parcels delivered in the Greater Toronto Area in the first half of 2017. We investigate different types of convolutional neural networks and demonstrate how our models outperform several baselines, from classical machine learning models to referenced solutions. Specifically, we show that a ResNet network with 8 residual blocks displays the best performance-complexity trade-off. We provide a thorough error analysis and visualize the features learned to better understand the model behavior, with remarks on data predictability. Our system has the potential to improve the user experience by better modeling their anticipation and to aid last-mile postal logistics as a whole.
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 | 0.000 | 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 | 0.001 |
| 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".