Development of a reinforcement learning-based adaptive scheduling algorithm for commercial smart kitchens
Bibliographic record
Abstract
Reinforcement learning (RL) is a machine learning method in which a model optimizes its decision-making strategy based on rewards or penalties received for the actions it takes in an environment, often simulated. An example of an optimized process could be work scheduling in a restaurant, with the cost function being the absolute error of the difference between the scheduled and actual delivery times of an order. In task planning, RL stands out for its ability to handle problems requiring a complex sequence of actions, where traditional planning algorithms may struggle. RL models can effectively explore the solution space, adjusting their decisions to changing conditions, which enables dynamic and adaptive task execution management. RL is a broad class encompassing various approaches to achieving a goal, and in this research, we focus on selected ones. Three popular RL methods named DQN, SARSA and TD-AC have been implemented and evaluated. The study was conducted in a simulated environment designed to replicate a "delivery-based" restaurant business model. The kitchen simulation model has been developed based on 65,845 recorded food preparation processes performed in 30 restaurants located throughout Poland. A rule-based, queue-driven model (FIFO) served as the baseline for absolute quality comparison of the generated schedules. The results show that, for the defined problem, the quality of the scheduling outcomes varies significantly depending on the choice of learning algorithm. Notably, the hybrid approach performed best under simulation conditions, considerably reducing the total completion time in a scenario reflecting the operations of a small, typical restaurant.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 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) | 0.000 | 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 teacher head, 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".