Situated and sequential planning and prediction of human driving behavior as decision making support system
DOI: 10.29007/6fxb
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper addresses the challenge of ensuring safe networked traffic by developing a decision-making support system capable of planning and predicting human driving behaviors. The primary motivation is to supervise human-machine interaction by detecting missing actions that could lead to critical situations, warning the driver, and potentially taking over driving functionality if necessary. The authors propose a situated and sequential planning approach that generates admissible action sequences in real-time, considering dynamic environmental changes and the predicted behaviors of surrounding traffic vehicles. The methodology combines Situation Operator Modeling (SOM) with trajectory prediction algorithms. SOM serves as an event-discrete approach where driving scenes are modeled as situations and driving actions as operators, forming a graph-based sequence from an initial to a desired final situation. To account for the behavior of other vehicles, the system employs a Long Short-Term Memory (LSTM) Encoder-Decoder algorithm with conventional social pooling layers to predict future trajectories. These predictions are integrated into a decision support control loop that calculates subsequent situations, checks assumptions against environmental conditions (such as Time to Collision thresholds), and evaluates goal reachability. The system prioritizes actions that achieve goals without triggering warnings and favors shorter operation times when safety conditions are equal. The approach was validated using the HighD dataset, which contains naturalistic vehicle trajectories recorded by drones on German highways. The specific application tested was an overtaking maneuver on a highway. The system defined performance conditions, including a critical speed of 130 km/h and specific Time to Collision (TTC) warning and conflict values for vehicles in adjacent lanes. In the simulation, the system evaluated multiple operators (acceleration, deceleration, waiting, steering) at each step. For instance, at the initial state, "acceleration" and "waiting" were viable options to reach a safe distance from the front vehicle. The system selected "waiting" because it provided a safer TTC margin compared to acceleration. Subsequent steps involved steering left into the free lane, steering right back into the original lane, and accelerating to complete the overtake. The algorithm successfully generated a sequential behavior plan (waiting, steering left, steering right, acceleration) that avoided conflicts and achieved the desired goal. The significance of this work lies in its demonstration of an automated supervisory control system that can continuously support human operators by predicting admissible actions and detecting potential errors in advance. By integrating trajectory prediction with graph-based situation modeling, the system offers a framework for real-time decision support that enhances safety in mixed human-autonomous traffic environments. The authors conclude that this methodology enables the detection of missing actions and the provision of alternative actions to reach desired situations, with future work aimed at validating the approach in real-time systems and expanding action-space based planning.
Key finding
The proposed Situation Operator Modeling combined with LSTM-based trajectory prediction successfully generates safe, sequential driving action plans for highway overtaking maneuvers by filtering out actions that violate environmental safety constraints.
Methodology
simulation_modeling
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-28 |
| archive | success | unpaywall | — | — | 2 | 2026-06-04 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | — | — | — | 1 | 2026-05-28 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
- anticipation
- situational awareness
- mental model of traffic
- automation surprise
- crash reconstruction hf
- gap acceptance
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Theoretical Contribution: computational model, theory or model, conceptual framework