BiLSTM Guided LPA Planning, Re-Planning, and Backtracking for Effective and Efficient Emergency Evacuation

1. Introduction

Indoor emergency evacuation presents a complex and highly dynamic decision-making problem. During a crisis, occupants must be guided from arbitrary locations to safe exits while hazards evolve rapidly and unpredictably, leveraging systems such as smart-building digital twins, real-time indoor positioning platforms, and intelligent sensor networks [1]. As illustrated in Figure 1, four major categories of challenges shape this problem: environmental hazards, information and sensing uncertainties, human and crowd dynamics, and algorithmic limitations. Environmental conditions can degrade within seconds—fires spread, smoke reduces visibility, debris blocks corridors, and structural failures render predefined routes unusable. Traditional evacuation approaches that rely on static floor plans or fixed procedures often become ineffective under such circumstances, resulting in costly re-planning or unsafe routing. A second challenge arises from incomplete and unreliable situational information. Indoor sensors may be sparse, malfunction during a disaster, or lack the communication bandwidth needed to provide timely updates on hazard evolution or exit accessibility. Evacuation algorithms must therefore operate under partial knowledge and adapt as new information becomes available—for instance, when a corridor becomes blocked or when a fire alarm signals a newly hazardous zone. Evacuation complexity is further amplified by human and crowd behaviour. Large groups tend to move unpredictably when stressed, forming congestion at bottlenecks, reversing direction abruptly, or deviating from suggested paths. These behavioural dynamics introduce significant uncertainty, complicating route prediction and degrading the effectiveness of rigid evacuation plans.

Existing computational methods address these challenges only partially. Rule-based systems encode fixed procedures but rarely adapt once conditions change. Simulation-based models such as computational fluid dynamics or agent-based simulations offer rich predictive power but are too computationally intensive for real-time use. Graph-based algorithms and linear programming methods compute optimal paths efficiently but typically assume static environments and lack semantic understanding of building spaces. More recently, ML! (ML!) approaches have been explored, yet they require large-scale annotated data and often generalize poorly to new or rapidly evolving scenarios unless continuously retrained.

A fundamental limitation across these methods is the absence of explicit semantic knowledge describing the structure and constraints of the building environment. Treating the space as a simple graph of nodes and edges ignores essential contextual information—for example, whether a location is a room or corridor, whether an area is restricted, or whether a region is likely to become hazardous. Without semantic context, planners may inadvertently choose routes that are unsafe or infeasible, such as directing evacuees into elevator lobbies misidentified as exits.

Building on recent insights in intelligent evacuation planning, this paper proposes a hybrid framework that integrates the computational efficiency of heuristic search with the adaptive intelligence of ontology-based reasoning. At its core, the proposed system employs LPA*! (LPA*!), an incremental variant of A*, to achieve fast and dynamic route planning by reusing prior search results as hazards evolve. A semantic reasoning module then continuously evaluates the ontology-informed state of the environment, detects when hazards invalidate the current plan, and triggers real-time adjustments or semantic backtracking to maintain safety and policy compliance. To support this adaptive behaviour, the proposed system leverages advances in semantic technologies, particularly those enabled by formal knowledge representation. Ontology engineering provides a structured, machine-interpretable model of the building environment—including rooms, corridors, stairwells, exits, and hazards—together with the relationships and constraints that govern their safe use during emergencies. By encoding this domain knowledge in both the OWL! (OWL!) and the RDF! (RDF!), the system benefits from automated reasoning that can infer implicit spatial connections, detect policy violations, and enhance the planner’s situational awareness. Unlike black-box ML! approaches, the ontology-driven layer in the proposed system offers transparency, semantic richness, and explainability. It can justify its decisions—such as excluding a route because it passes through a restricted or unsafe area—thereby improving trustworthiness and supporting accountable, safety-compliant evacuation management. In summary, this paper makes the following original contributions:

• Ontology-Driven Building Model: We propose a formal OWL!/RDF!-based ontology to represent both the spatial structure and semantic constraints of a multi-floor building environment, including rooms, corridors, staircases, exits, hazards, and evacuation policies. This knowledge layer enables automated reasoning over implicit connections and safety rules during evacuation planning.
• Reasoning-Enhanced Heuristic for LPA*: We introduce a semantic heuristic function that exploits ontology-derived information to guide the LPA*! algorithm. By integrating inferred paths, vertical navigation constraints, and policy-compliant transitions, the planner achieves efficient incremental re-planning without full recomputation.
• BiLSTM-Guided Neuro-Symbolic Planning: We develop a neuro-symbolic extension of semantic LPA*! by incorporating a BiLSTM! (BiLSTM!) model trained on ontology-consistent evacuation trajectories. The BiLSTM! predicts remaining evacuation cost and the likelihood of backtracking, providing anticipatory guidance while preserving admissibility through a bounded hybrid heuristic.
• Dynamic Re-planning and Semantic Backtracking Mechanism: We design an event-driven re-planning and backtracking strategy that allows the planner to revise unsafe routes in response to evolving hazards (e.g., blocked corridors or structural failures) and to safely return to the last feasible semantic situation when necessary.
• Explainable and Policy-Compliant Decision Making: Unlike black-box learning approaches, the proposed framework provides explainable evacuation decisions grounded in ontology reasoning and SWRL rules, ensuring that generated routes comply with safety policies and access constraints.
• Realistic Multi-Floor Case Study: We provide a detailed case study conducted in a realistic multi-floor academic building to demonstrate the applicability of the proposed framework under dynamic evacuation scenarios.

The remainder of this paper is organised as follows. Section 2 reviews related work and identifies the research gaps addressed in this study. Section 3 presents the theoretical background, including LSTM! (LSTM!) and BiLSTM! models for temporal prediction and backtracking in evacuation planning. Section 4 details the proposed neuro-symbolic framework, including the ontology-driven building model, semantic reasoning layer, reasoning-enhanced LPA*! heuristic, BiLSTM!-guided hybrid heuristic, and dynamic re-planning and backtracking mechanisms. Section 5 reports the experimental setup and quantitative results, including ablation studies comparing semantic LPA*! and BiLSTM!-guided variants. Finally, Section 6 concludes the paper and outlines limitations and future research directions.

2. Related work

Several studies have been proposed in the literature to address emergency evacuation planning, spanning classical optimisation models, heuristic algorithms, and simulation-based approaches. Existing contributions differ significantly in terms of spatial scale, uncertainty modelling, and the extent to which risk, routing, and resource coordination are jointly considered. To contextualise the proposed approach, prior works are reviewed below according to the scale of the evacuation environment.

• Outdoor and City-Scale Evacuation Planning: Large-scale evacuation planning has traditionally focused on network-level optimisation under uncertainty. Goerigk and Grün [2] formulated a robust bus evacuation problem that explicitly accounts for delayed information on evacuee demand, demonstrating how robust optimisation can mitigate uncertainty in urban-scale evacuations involving transit-dependent populations. While effective at the scheduling level, this model does not explicitly incorporate spatially heterogeneous risk or route-level exposure.
  To address congestion and route diversity, Chang et al. [3] studied the problem of identifying k discriminative evacuation paths on road networks, aiming to minimise both travel distance and path overlap. Their ant colony optimisation-based heuristic enables congestion mitigation through path diversification, which is particularly relevant for emergency evacuation and rescue logistics. However, environmental hazards and spatial risk intensity are not explicitly modelled [4].
  More recent studies have integrated environmental dynamics into city-scale evacuation planning. Li et al. [5] proposed a coupled hydrodynamic and CA! (CA!)-based framework for pedestrian evacuation under flooding scenarios, explicitly accounting for flood depth, flow velocity, and human instability. Similarly, Shao et al. [6] emphasised exposure-based evacuation modelling, highlighting the importance of spatially varying hazard intensity in flood-prone urban regions.
  Uncertainty-aware urban evacuation routing has also been investigated using heuristic optimisation techniques. Mao et al. [7] developed a fuzzy integer programming model combined with an improved ant colony optimisation algorithm to minimise expected evacuation time in uncertain traffic environments, addressing both symmetric and asymmetric evacuation demands. Although computationally efficient, the approach remains primarily route-centric and does not consider coordinated decisions on shelters or relief distribution. More recently, Jafarian et al. [8] and Li et al. [9] investigated large-scale evacuation from complementary perspectives. Jafarian et al. [8] focused on the impact of behavioural factors on evacuation efficiency, analysing how individual compliance, response delays, and behavioural heterogeneity influence route choice and overall evacuation performance under emergency conditions. Their work enhances behavioural realism but assumes a fixed evacuation network and does not explicitly optimise risk distribution across geographical zones. In contrast, Li et al. [9] adopted a collaborative evacuation framework in which multiple stakeholders or decision-making entities coordinate evacuation actions to improve system-level efficiency. While this approach captures inter-agent coordination and information sharing, geographical risk is still embedded indirectly through network parameters rather than being modelled as an explicit optimisation criterion. Consequently, neither approach directly addresses the problem of balancing spatially heterogeneous risk across the evacuation network.
• Single-Floor Building Evacuations and Simulations: At the building scale, evacuation research has largely relied on simulation-based and heuristic path-planning approaches. Wang et al. [10] proposed a cellular ant optimisation model for passenger ship evacuation, combining CA! with ant colony optimisation on hexagonal grids to accelerate search and reduce the likelihood of local optima. While effective in confined environments, the method assumes static hazards and does not explicitly account for dynamically evolving risk.
  Infrastructure-assisted evacuation has also been investigated. Zhang et al. [11] addressed evacuation efficiency through the optimal placement of signage systems in public spaces, modelling pedestrian–signage interactions using a cooperative location framework. Although this improves wayfinding and guidance, the approach focuses on behavioural support rather than adaptive evacuation routing under emergency conditions.
  More recently, Baglioni and Jamshidnejad [12] introduced a model predictive control framework for indoor search-and-rescue robots, integrating target-oriented and coverage-oriented objectives under uncertainty. While their approach demonstrates strong performance in dynamic indoor environments, it is primarily designed for robotic exploration and does not directly address large-scale human evacuation or network-level risk balancing.
• Multi-Floor and Complex Building Evacuation: Evacuation in complex and multi-level structures introduces additional challenges related to vertical movement, smoke propagation, and inter-floor connectivity. Li and Huang [13] proposed a 3D GIS-based evacuation framework for large public buildings that integrates numerical fire simulation, smoke diffusion, and individual behaviour modelling with A* path planning. Their approach enables risk-aware routing across multiple floors, but relies heavily on simulation outputs and lacks an explicit optimisation mechanism for balancing risk across the evacuation network.
  In specialised underground environments, Zheng et al. [14] investigated coordinated rescue-path planning for multiple mine rescue teams using a hybrid FA-MDPSO algorithm and force-directed graph layouts. Their work highlights the importance of multi-team coordination and network visualisation in constrained environments. However, the proposed framework is domain-specific and does not readily generalise to urban evacuation networks involving heterogeneous geographical risk and multiple decision layers.

Across the surveyed literature, most evacuation planners rely either on optimisation (e.g., mixed-integer linear programming (MILP), fuzzy integer programming (IP)), meta-heuristics (ACO! (ACO!), PSO! (PSO!) variants), or agent-based and CA!-based simulation. As Table 1 shows, these systems typically operate on raw grid or graph abstractions, with limited or no semantic understanding of the built environment; almost none incorporate ontological reasoning. Likewise, key indoor-evacuation requirements—such as multi-floor navigation, dynamic re-planning, backtracking, vertical movement, and real-time event adaptation—are only partially supported and often absent. Explainability is also consistently missing, with all prior studies lacking XAI capabilities.

Our work positions itself at this intersection of unresolved issues. Instead of treating buildings as unstructured spaces, this paper introduces a formal building ontology that encodes the semantics of rooms, corridors, exits, constraints, and access policies. This knowledge layer fills the interpretability gap noted in prior work and enables the planner to enforce high-level rules (e.g., avoid hazardous spaces or restricted “Staff Only’’ areas). Combined with a dynamic heuristic graph search method (LPA*!), our system supports continuous re-planning, multi-floor movement, backtracking, mass evacuation, and real-time adaptation—capabilities that prior studies either lack or offer only partially. The result is a hybrid, knowledge-aware planner that delivers explainable decisions while maintaining the efficiency of algorithmic search. To the best of our knowledge, this semantic–algorithmic integration is novel in the context of indoor evacuation planning.

3. Preliminaries

Effective emergency evacuation requires dynamic planning capabilities that can account for evolving conditions such as congestion, blocked paths, and time-dependent hazards. Sequence learning models, particularly LSTM! networks and their bidirectional variant (BiLSTM!), provide a powerful framework for capturing temporal dependencies in evacuation-related data, enabling both forward planning and informed backtracking decisions.

3.1. LSTM for Evacuation Planning

In evacuation planning, the objective is to predict future system states (situations)—such as crowd density, travel time, or route feasibility—based on historical and real-time observations. LSTM! networks are well suited for this task due to their gated memory structure, which allows the model to retain critical past information while discarding irrelevant dynamics.

At time step t, given an input vector ${\mathbf{x}}_t$ representing evacuation features (e.g., node occupancy, edge capacity, hazard level), the LSTM! updates its internal states as Equation 1:

$$\left\{\begin{array}{cc}\hfill {\mathbf{f}}_t& =\sigma ({\mathbf{W}}_f[{\mathbf{h}}_{t-1},{\mathbf{x}}_t]+{\mathbf{b}}_f),\hfill \\ \hfill {\mathbf{i}}_t& =\sigma ({\mathbf{W}}_i[{\mathbf{h}}_{t-1},{\mathbf{x}}_t]+{\mathbf{b}}_i),\hfill \\ \hfill {\tilde{\mathbf{c}}}_t& =tanh({\mathbf{W}}_c[{\mathbf{h}}_{t-1},{\mathbf{x}}_t]+{\mathbf{b}}_c),\hfill \\ \hfill {\mathbf{c}}_t& ={\mathbf{f}}_t\odot {\mathbf{c}}_{t-1}+{\mathbf{i}}_t\odot {\tilde{\mathbf{c}}}_t,\hfill \\ \hfill {\mathbf{o}}_t& =\sigma ({\mathbf{W}}_o[{\mathbf{h}}_{t-1},{\mathbf{x}}_t]+{\mathbf{b}}_o),\hfill \\ \hfill {\mathbf{h}}_t& ={\mathbf{o}}_t\odot tanh\left({\mathbf{c}}_t\right),\hfill \end{array}\right.$$

(1)

where ${\mathbf{h}}_t$ encodes a predictive representation of the evolving evacuation state. This forward temporal modeling supports proactive planning by anticipating congestion propagation and estimating future route viability.

3.2. BiLSTM for Backtracking and Route Revision

Emergency evacuation often requires revisiting prior decisions when unexpected events occur, such as sudden bottlenecks or hazard escalation. BiLSTM! networks enhance decision-making by integrating both past observations and future contextual information within a planning horizon.

A BiLSTM! layer consists of a forward LSTM! that processes the evacuation sequence in chronological order and a backward LSTM! that processes it in reverse order. The forward hidden state ${\overrightarrow{\mathbf{h}}}_t$ captures planning-oriented dynamics, while the backward hidden state ${\overleftarrow{\mathbf{h}}}_t$ supports backtracking by evaluating how future constraints affect earlier routing choices. The combined representation is given by:

$${\mathbf{h}}_t=[{\overrightarrow{\mathbf{h}}}_t\left|\right|{\overleftarrow{\mathbf{h}}}_t].$$

(2)

By jointly modelling forward planning and backwards feasibility, BiLSTM! layers provide a unified representation that enables effective and efficient evacuation strategies, supporting adaptive rerouting, decision revision, and resilience under uncertainty.

4. Research Methodology

This section describes the methodology of the proposed heuristic and knowledge-based evacuation planning system supported by a BiLSTM! prediction module. As shown in Figure 2, the architecture combines symbolic semantic reasoning with data-driven learning to support adaptive evacuation planning and is organised around micro-models capturing topology, events, situations (states), and actions, a planning layer encoding prescriptive, descriptive, and policy rules, and a controller responsible for rule interpretation, heuristic evaluation, and event processing. The detailed structure and composition of each component are discussed in subsequent sections.

To make the neuro-symbolic stack explicit, Table 2 summarises what each layer consumes and produces, and how the responsibilities compose at run time. In execution, events are asserted into the ontology, reasoning and SWRL constraints refresh the set of admissible transitions, and the planner searches over semantic situations using an admissible semantic heuristic. If an event changes costs or invalidates a transition, the replanner incrementally repairs the current plan rather than recomputing from scratch; if a transition fails at action execution time, the backtracker rolls back to the most recent feasible semantic situation and triggers replanning from that point. The BiLSTM module provides guidance signals that bias search towards stable trajectories and more importantly, it does not introduce new actions and does not override SWRL admissibility: it only biases the ordering of expansions through the bounded heuristic in Eq. (9), while SWRL rules and constraints govern which transitions are permitted.

Because evacuation unfolds under randomly triggered events, the planner cannot rely on a static, offline distance map that assumes a single fixed best route. Grid-based distance heuristics (often instantiated as Manhattan distance on rectilinear floor discretisations) are commonly used to approximate corridor traversal in indoor route computation [15,16], but they remain purely geometric and do not encode evacuation policies, restricted areas, or event-driven hazards that can invalidate a route at run time. In our approach, the building state is represented in the ontology, and evacuation behaviour is formulated as policy-driven transitions expressed as (state, action, next-state) triplets. When an event occurs (for example, a fire outbreak [16], congestion build-up [17], or a blocked corridor [16]), the re-planner queries the updated knowledge base to revise the admissible actions and regenerate a feasible action sequence to the goal state. The BiLSTM prediction module complements this symbolic layer by providing temporal guidance on likely congestion evolution from prior trajectories, enabling earlier route revision and backtracking where needed; however, all candidate moves are still filtered by the ontology-defined (state, action, next-state) transitions and the current event assertions, so only semantically admissible actions are expanded. Moreover, admissible moves are instantiated and validated via SWRL! (SWRL!) triplets, ensuring that each expansion corresponds to a policy-compliant (state, action, next-state) transition.

Figure 3 summarises three representative evacuation sequences (state–action sequences); corrective actions (CA) indicate rerouting, and backtracking actions (BTA) indicate deliberate returns to earlier decision points, while each step is an admissible (state, action, next-state) transition instantiated from the SWRL triplets in Table 3 under the current event assertions.

4.1. Evacuation Micro-Models

Evacuation micro-models provide the atomic semantic representations required by the proposed planning and learning algorithms. These models define how evacuees, environments, and actions are represented at a fine-grained level, forming the basis for ontology reasoning, heuristic evaluation, and state-space search.

• Topological Modelling: The evacuation environment is represented as a directed semantic graph $\mathcal{G}=(\mathcal{V},\mathcal{E})$, where each node $v\in \mathcal{V}$ corresponds to a physical space (e.g., room, corridor, staircase, exit), and each edge $e\in \mathcal{E}$ represents a traversable connexion. This topological structure enables the retrieval of explicit neighbours during planning, as follows (Equation 3):

  $$\mathtt{QueryNeighbors}(\mathtt{current},\mathcal{O}).$$

  (3)
• Event Modeling: Events represent discrete occurrences that may alter evacuation dynamics, such as door closures, congestion, alarms, or accessibility changes. Events are represented as assertions that update the ontology’s state, thereby altering the availability of transitions or the cost during planning. Formally, an event $e_t$ induces a change in the transition function (Equation 4):

  $$(s_t,a_t)\to s_{t+1},$$

  (4)

  which is dynamically reflected in the inferred neighbour set queried by the planner.
• Situation Representation: A situation corresponds to a semantic state describing the current evacuation context of an individual. Each situation $s\in \mathcal{S}$ encapsulates: (i) the evacuee’s location, (ii) the structural properties of the space, and (iii) safety and accessibility constraints. Situations serve as nodes in the state-space search graph explored by the LPA*! algorithm.
• Action Space Definition: Actions define admissible movements between situations, such as move_to_corridor, enter_staircase, or exit_building. The action space $\mathcal{A}$ is constrained by ontology rules and encoded using SWRL! triplets. Each action induces a cost $c(s,a,s^{\prime })$ used during planning (Equation 5):

  $$c=Cost(\mathrm{current},\mathrm{neighbor}),$$

  (5)
  These costs may reflect distance, vertical movement penalties, or safety-related preferences.

4.2. Planning Rules

Planning rules encode domain knowledge that governs admissible evacuation behaviour. These rules are formalised using SWRL! and operate over the ontology to define valid transitions, constraints, and preferences during evacuation planning.

• Prescriptive Rules: Prescriptive rules define mandatory evacuation actions that must be taken under specific conditions, such as enforcing movement toward exits once an alarm is triggered or preventing access to restricted areas. These rules ensure compliance with evacuation policies and safety regulations.
  Formally, prescriptive rules constrain the action space by allowing only transitions that satisfy predefined safety conditions:

  $$(s_t,a_t)\to s_{t+1}\mathrm{if}\mathrm{and}\mathrm{only}\mathrm{if}{\mathcal{R}}_{\mathrm{pres}}\left(s_t\right)=\mathrm{true}.$$
• Descriptive Rules: Descriptive rules capture factual relationships within the environment, such as adjacency between spaces, floor connectivity, and accessibility relations. These rules enable the ontology reasoner to infer implicit connections that are not explicitly encoded in the topological graph. Such inferred relations are queried during planning through ontology reasoning, as reflected in Algorithm 3 by the function:

  $$\mathtt{QueryInferredConnections}(\mathtt{current},\mathcal{O}).$$
• Policy Rules: Policy rules express evacuation preferences rather than strict constraints. Examples include prioritising safer routes, avoiding congested areas, or favouring ramps over stairs for mobility-impaired evacuees. These rules influence the planning process by affecting heuristic evaluation and cost assignment rather than enforcing hard exclusions.
• Constraint Encoding: Constraint rules encode physical and operational limitations, such as blocked passages, closed doors, or capacity restrictions. These constraints dynamically prune the state–transition graph by removing infeasible neighbours during node expansion.

Together, planning rules ensure that all paths explored by the planner are semantically valid, policy-compliant, and context-aware while remaining flexible to dynamic environmental changes.

4.3. Controller

The controller constitutes the operational core of the proposed evacuation framework and integrates the functionalities previously described in the system architecture, namely the ontology and reasoning engine, heuristic path planner, event generator, and action interpreter, into a unified execution loop.

At its foundation, the controller maintains an OWL 2 ontology representing the building layout and evacuation knowledge, including spatial entities (e.g., Room, Corridor, Exit), dynamic entities (e.g., Person, Hazard), and action concepts (e.g., EvacuationAction, BlockedPath). Safety policies and access constraints are encoded using axioms and SWRL! rules so that inadmissible transitions are filtered during planning.

Dynamic events generated by the simulator (such as smoke propagation or blocked corridors) are translated into ontology assertions. These updates trigger the reasoning process, which refreshes inferred relations (e.g., admissible connectivity or safety classifications) and signals the planner that replanning is required.

• Rule Engine: Applies SWRL!-based planning rules to validate actions and infer implicit state transitions. Routes violating evacuation policies (e.g., restricted areas or unsafe corridors) are rejected at expansion time.
• Semantic Interpreter: Maps ontology assertions and inferred relations into executable planning constructs. Semantic situations correspond to graph nodes, actions to transitions, and inferred relations to neighbour sets used by the planner.
• Heuristic Engine: Implements the heuristic path planner using semantic LPA*!. At each expansion step, both explicit neighbours (physical adjacency) and inferred neighbours (reasoner-derived connections) are evaluated using the semantic heuristic $h_o\left(n\right)$ and, when enabled, the learned guidance $h_l\left(n\right)$:

  $$f\left(n\right)=g\left(n\right)+h\left(n\right).$$
• Event Processing Mechanism: Produces and processes time-stamped dynamic events (e.g., hazards, blocked corridors, door constraints) and updates the ontology accordingly. These updates modify edge costs and admissibility, which triggers incremental replanning.
• Action Interpreter: Translates the planner’s sequence of situations and actions into executable evacuation guidance. Each transition is validated against the current ontology state. If a transition becomes invalid due to a new event, semantic backtracking is invoked and replanning resumes from the last feasible situation.

Because semantic events usually affect only a limited subset of edges and constraints, the controller performs incremental repair rather than full recomputation. Only affected vertices and their local neighbourhood are updated and propagated through the priority queue, preserving consistency with updated semantic constraints while maintaining efficiency.

Algorithm 1 formalises this closed-loop control process, including event handling, replanning triggers, and action execution.

Algorithm 1Dynamic Path Planning with Re-planning and Backtracking

Require: InitialSituation, OccupantID, BuildingTopology, EventDefinitions, ActionRules Ensure: Safe exit notification or reroute guidance 1: procedure  InitializeSystem 2:     $current\leftarrow FETCHINITIALSITUATION$ 3:     BuildTopologyModel(BuildingTopology) 4:     $eventDefs\leftarrow \mathrm{LOADEVENTDEFINITIONS}$ 5:     $actionRules\leftarrow \mathrm{LOADACTIONRULES}$ 6:     InitializeTester 7: end procedure 8: procedure  MainLoop 9:     InitializeSystem 10:     while not ExitReached(OccupantID) do 11:         $event\leftarrow \mathrm{FETCHNEXTEVENT}$ 12:         $type\leftarrow \mathrm{CLASSIFYEVENT}(event,eventDefs)$ 13:         if $type=\text{“}\mathrm{Impact}\text{”}$ then 14:            HandleImpactEvent(current, event) 15:         else 16:            ContinueNormalPlan(current) 17:         end if 18:         if TesterCheckFails(current) then 19:            $trigger\leftarrow \mathrm{IDENTIFYTRIGGER}\left(current\right)$ 20:            $pla{n}^{\prime }\leftarrow \mathrm{REPLAN}(trigger,actionRules)$ 21:            UpdateGuidance(OccupantID, plan’) 22:            $current\leftarrow \mathrm{UPDATESITUATION}(OccupantID,pla{n}^{\prime })$ 23:         end if 24:         $current\leftarrow \mathrm{FETCHNEXTSITUATION}$ 25:     end while 26:     ConfirmExit(OccupantID) 27: end procedure 28: procedure HandleImpactEvent($situation,event$) 29:     $route\leftarrow \mathrm{FINDSAFEPATH}(situation,event,actionRules)$ 30:     RelayRoute(OccupantID, route) 31:     LogEventAndRoute(event, route) 32: end procedure 33: procedure ContinueNormalPlan($situation$) 34:     $action\leftarrow \mathrm{SELECTACTION}(situation,actionRules)$ 35:     ExecuteAction(action) 36:     LogExecution(action, situation) 37: end procedure 38: procedure TesterCheckFails($situation$) 39:     $expected\leftarrow \mathrm{PREDICTOUTCOME}\left(situation\right)$ 40:     $actual\leftarrow \mathrm{FETCHACTUALSITUATION}$ 41:     return $actual\ne expected$ 42: end procedure 43: procedure IdentifyTrigger($situation$) 44:     $lastEvent\leftarrow \mathrm{FETCHLASTEVENT}$ 45:     return AnalyzeImpact(lastEvent, situation) 46: end procedure

4.4. Neuro-Symbolic Evacuation Planning Framework

This subsection formalises the neuro-symbolic planning methodology that combines ontology-driven semantic reasoning, heuristic graph search, and BiLSTM-based temporal guidance. It operationalises the components described in the controller through three interacting layers: semantic heuristic estimation, incremental planning, and learning-guided refinement.

•

Reasoning-Enhanced Semantic Heuristic: The semantic heuristic $h_o\left(n\right)$ exploits ontology-derived structure that cannot be captured by purely geometric heuristics. It queries inferred evacuation paths, vertical navigation requirements, and policy-compliant transitions to estimate remaining cost:

$$h_o\left(n\right)=\left\{\begin{array}{cc}\mathrm{GeometricHeuristic}(n,g),\hfill & \mathrm{if}\mathrm{no}\mathrm{inferred}\mathrm{path}\mathrm{exists},\hfill \\ (\mathrm{horizontalCost}+\mathrm{verticalCost})\times \mathrm{qualityFactor},\hfill & \mathrm{otherwise}.\hfill \end{array}\right.$$

(6)

When a plausible route can be inferred by the reasoner, the heuristic converts semantic evidence into an optimistic estimate combining horizontal traversal and vertical transitions. The factor $\mathrm{qualityFactor}(n,g)\in (0,1]$ acts as a confidence discount so that uncertain inferred routes are down-weighted without violating admissibility. Algorithm 2 specifies this computation.

Algorithm 2 Heuristic Estimation Using Inferred Paths

Require: current node n, goal node g Ensure: heuristic estimate $h_o\left(n\right)$ is returned 1: $\mathrm{paths}\leftarrow \mathrm{QUERYINFERREDPATHS}(n,g)$▹ Candidate inferred route sequences from n to g via ontology reasoning. 2: $\mathrm{h\_geo}\leftarrow \mathrm{GEOMETRICHEURISTIC}(n,g)$▹ Admissible geometric fallback (e.g., Manhattan/Euclidean in the discretisation). 3: if$\mathrm{ paths}= \varnothing$ then 4:     return $\mathrm{h\_geo}$ 5: end if 6: $\mathrm{minHops}\leftarrow min\left\{\right|p|:p\in \mathrm{paths}\}$▹ Length of shortest inferred path in number of hops. 7: $\mathrm{avgCost}\leftarrow \mathrm{ESTIMATEAVERAGEEDGECOST}\left(n\right)$▹ Optimistic average traversal cost from this node per edge. 8: $\mathrm{currentFloor}\leftarrow \mathrm{GETFLOOR}\left(n\right)$ 9: $\mathrm{goalFloor}\leftarrow \mathrm{GETFLOOR}\left(g\right)$ 10: $\mathrm{floorDiff}\leftarrow \left|\mathrm{currentFloor}-\mathrm{goalFloor}\right|$ 11: $\mathrm{verticalCost}\leftarrow \mathrm{floorDiff}\times \left({\displaystyle \frac{\mathrm{FloorHeight}}{\mathrm{ClimbSpeed}}}\right)$▹ Optimistic time to change floorDiff floors via stairs. 12: $\mathrm{horizontalCost}\leftarrow \mathrm{minHops}\times \mathrm{avgCost}$ 13: $\mathrm{quality}\leftarrow \mathrm{ASSESSPATHQUALITY}\left(\mathrm{paths}\right[0\left]\right)$▹ Assess inferred path plausibility under current hazards/policies. 14: $\mathrm{qFactor}\leftarrow \mathrm{QUALITYFACTOR}\left(\mathrm{quality}\right)$▹$\mathrm{qFactor}\in (0,1]$ is a confidence discount: 1 if high-quality, smaller if uncertain. 15: $\mathrm{h\_inf}\leftarrow \left(\mathrm{horizontalCost}+\mathrm{verticalCost}\right)\times \mathrm{qFactor}$ 16: return$max(\mathrm{h\_geo},\mathrm{h\_inf})$▹ Max of admissible heuristics is admissible and more informative.

•

Semantic Lifelong Planning A*: Semantic LPA*! integrates the heuristic $h_o\left(n\right)$ into an incremental planning framework. The search space is defined by semantic states derived from the ontology, and neighbour expansion combines explicit topological connections with inferred semantic transitions. For each node n, the planner evaluates:

$$f\left(n\right)=g\left(n\right)+h_o\left(n\right).$$

(7)

When the ontology is updated due to dynamic events (e.g., blocked corridors or modified traversal costs), only the affected nodes and edges are updated, and vertex consistency is restored using the standard LPA*! bookkeeping (g- and $\mathit{rhs}$-values). This enables rapid local repair instead of full recomputation.

•

Neuro-Symbolic Semantic LPA* (BiLSTM-Guided): To incorporate temporal experience, a BiLSTM model is trained offline on SWRL!-consistent state-action sequences expressed in the same symbolic vocabulary as the planner. At run time, given the movement history $\mathcal{H}$ and a candidate node n, the model predicts a guidance term:

$$h_l\left(n\right)=\mathrm{BiLSTM}(\mathcal{H}\cup \left\{n\right\}).$$

(8)

• BiLSTM-guided bounded heuristic.

Let $h_o\left(n\right)$ denote the ontology-derived admissible heuristic estimate of the remaining cost to the goal. The BiLSTM outputs a learned remaining-cost estimate $h_l\left(n\right)\ge 0$ from the recent state–action history. We combine both signals using a bounded convex blend:

$$h^{\prime }\left(n\right)=min\left(h_o\left(n\right),(1-\lambda )h_o\left(n\right)+\lambda h_l\left(n\right)\right),0\le \lambda \le 1.$$

(9)

By construction $h^{\prime }\left(n\right)\le h_o\left(n\right)$ for all n. Since $h_o$ is admissible, $h^{\prime }$ remains admissible, hence the optimality guarantees of semantic LPA*! are preserved. The parameter $\lambda$ controls the strength of the learned guidance whenever $h_l\left(n\right)<h_o\left(n\right)$; otherwise the heuristic falls back to $h_o\left(n\right)$.

Setting $\lambda =0$ recovers pure semantic LPA*!, while larger values increase the influence of learned guidance without violating optimality or policy constraints. The evaluation function becomes:

$$f\left(n\right)=g\left(n\right)+h^{\prime }\left(n\right).$$

(10)

Because $h^{\prime }\left(n\right)\le h_o\left(n\right)\le h^{*}\left(n\right)$, admissibility and optimality guarantees of LPA*! are preserved. The planner therefore benefits from temporal awareness—such as early detection of routes prone to backtracking—while maintaining explainability and semantic compliance. Algorithm 3 implements semantic LPA*! planning process along with incremental neuro-symbolic extension (BiLSTM-guided).