--- title: "Building a Pool Service Agent: Maintenance Scheduling, Chemical Balance, and Equipment Repair" description: "Build an AI agent for pool service companies that optimizes service routes, calculates chemical dosages, diagnoses equipment issues, and manages seasonal opening and closing schedules." canonical: https://callsphere.ai/blog/building-pool-service-agent-maintenance-chemical-balance-repair category: "Learn Agentic AI" tags: ["Pool Service", "Chemical Balance", "Service Routes", "Equipment Diagnostics", "Seasonal Planning"] author: "CallSphere Team" published: 2026-03-17T00:00:00.000Z updated: 2026-05-06T01:02:44.700Z --- # Building a Pool Service Agent: Maintenance Scheduling, Chemical Balance, and Equipment Repair > Build an AI agent for pool service companies that optimizes service routes, calculates chemical dosages, diagnoses equipment issues, and manages seasonal opening and closing schedules. ## The Pool Service Operations Model Pool service companies run route-based businesses. A technician visits 8-12 pools per day, testing water chemistry, adding chemicals, cleaning filters, and inspecting equipment. The difference between a profitable pool service company and a struggling one often comes down to route efficiency and chemical accuracy. An AI agent that optimizes routes, calculates exact chemical dosages, diagnoses equipment problems before they become emergencies, and manages seasonal transitions can increase the number of pools each technician services by 20-30%. Chemical balance is where the AI adds the most technical value. Pool chemistry involves multiple interacting variables — pH, alkalinity, calcium hardness, cyanuric acid, and sanitizer levels — where adjusting one affects the others. ## Chemical Balance Calculator Pool chemistry requires precise calculations based on pool volume, current readings, and target ranges. The agent calculates exact dosages. ```mermaid flowchart LR INPUT(["User intent"]) PARSE["Parse plus classify"] PLAN["Plan and tool selection"] AGENT["Agent loop LLM plus tools"] GUARD{"Guardrails and policy"} EXEC["Execute and verify result"] OBS[("Trace and metrics")] OUT(["Outcome plus next action"]) INPUT --> PARSE --> PLAN --> AGENT --> GUARD GUARD -->|Pass| EXEC --> OUT GUARD -->|Fail| AGENT AGENT --> OBS style AGENT fill:#4f46e5,stroke:#4338ca,color:#fff style GUARD fill:#f59e0b,stroke:#d97706,color:#1f2937 style OBS fill:#ede9fe,stroke:#7c3aed,color:#1e1b4b style OUT fill:#059669,stroke:#047857,color:#fff ``` ```python from dataclasses import dataclass from typing import Optional @dataclass class WaterTestResults: ph: float free_chlorine: float # ppm total_alkalinity: float # ppm calcium_hardness: float # ppm cyanuric_acid: float # ppm total_dissolved_solids: float # ppm temperature_f: float pool_volume_gallons: int TARGET_RANGES = { "ph": (7.2, 7.6), "free_chlorine": (1.0, 3.0), "total_alkalinity": (80, 120), "calcium_hardness": (200, 400), "cyanuric_acid": (30, 50), } class ChemicalCalculator: def calculate_adjustments(self, readings: WaterTestResults) -> list[dict]: adjustments = [] volume = readings.pool_volume_gallons # pH adjustment if readings.ph TARGET_RANGES["ph"][1]: excess = readings.ph - TARGET_RANGES["ph"][1] muriatic_oz = excess * volume / 10000 * 16 adjustments.append({ "parameter": "pH (lower)", "current": readings.ph, "target": TARGET_RANGES["ph"][1], "chemical": "Muriatic Acid (31.45%)", "amount_oz": round(muriatic_oz, 1), "instruction": "Add slowly to deep end with pump running. Retest in 4 hours.", }) # Chlorine adjustment if readings.free_chlorine dict: """Langelier Saturation Index: predicts scaling or corrosion tendency.""" import math temp_factor = 0.0 + (readings.temperature_f - 32) * 0.01 tf = round(temp_factor, 2) cf = round(math.log10(readings.calcium_hardness) - 0.4, 2) af = round(math.log10(readings.total_alkalinity), 2) lsi = readings.ph - (9.3 + tf + cf + af) lsi = round(lsi, 2) if lsi > 0.3: condition = "scaling" action = "Lower pH or calcium hardness to prevent scale buildup" elif lsi float: R = 3959 dlat = radians(lat2 - lat1) dlon = radians(lon2 - lon1) a = sin(dlat/2)**2 + cos(radians(lat1)) * cos(radians(lat2)) * sin(dlon/2)**2 return R * 2 * atan2(sqrt(a), sqrt(1 - a)) class RouteOptimizer: def optimize_daily_route( self, start_location: tuple, pools: list[dict], ) -> list[dict]: """Nearest-neighbor heuristic for route optimization.""" remaining = list(pools) route = [] current_lat, current_lon = start_location while remaining: nearest = min( remaining, key=lambda p: haversine(current_lat, current_lon, p["lat"], p["lon"]), ) distance = haversine(current_lat, current_lon, nearest["lat"], nearest["lon"]) route.append({ "stop": len(route) + 1, "address": nearest["address"], "customer": nearest["customer_name"], "distance_from_previous": round(distance, 1), "estimated_service_time_min": nearest.get("service_time", 30), "special_notes": nearest.get("notes", ""), }) current_lat, current_lon = nearest["lat"], nearest["lon"] remaining.remove(nearest) total_distance = sum(s["distance_from_previous"] for s in route) total_time = sum(s["estimated_service_time_min"] for s in route) return { "stops": route, "total_distance_miles": round(total_distance, 1), "total_service_time_hours": round(total_time / 60, 1), "estimated_drive_time_hours": round(total_distance / 25, 1), } ``` ## Equipment Diagnostics Pool equipment fails in predictable patterns. The agent diagnoses issues from symptoms and recommends repairs. ```python EQUIPMENT_DIAGNOSTICS = { "pump_not_priming": { "symptoms": ["pump running but no water flow", "air bubbles in pump basket"], "probable_causes": [ {"cause": "Air leak in suction line", "likelihood": "high", "fix": "Check and replace O-rings on pump lid and unions", "cost_range": "$15-45"}, {"cause": "Clogged impeller", "likelihood": "medium", "fix": "Remove pump housing and clear debris from impeller", "cost_range": "$85-150"}, {"cause": "Low water level", "likelihood": "high", "fix": "Fill pool to mid-skimmer level", "cost_range": "$0"}, ], }, "heater_not_firing": { "symptoms": ["heater turns on but no heat", "error codes on display"], "probable_causes": [ {"cause": "Dirty or failed pressure switch", "likelihood": "high", "fix": "Clean or replace pressure switch", "cost_range": "$45-120"}, {"cause": "Failed ignitor", "likelihood": "medium", "fix": "Replace hot surface ignitor", "cost_range": "$80-200"}, {"cause": "Low gas pressure", "likelihood": "low", "fix": "Contact gas company to check supply pressure", "cost_range": "$0"}, ], }, "filter_pressure_high": { "symptoms": ["pressure gauge above 25 PSI", "reduced water flow"], "probable_causes": [ {"cause": "Dirty filter cartridge or grids", "likelihood": "high", "fix": "Clean or replace filter media", "cost_range": "$0-300"}, {"cause": "Clogged return lines", "likelihood": "low", "fix": "Professional line cleaning required", "cost_range": "$150-350"}, ], }, } def diagnose_equipment(symptom_description: str) -> dict: description_lower = symptom_description.lower() for issue_key, issue in EQUIPMENT_DIAGNOSTICS.items(): for symptom in issue["symptoms"]: if any(word in description_lower for word in symptom.split()): return { "issue": issue_key.replace("_", " ").title(), "matching_symptoms": issue["symptoms"], "probable_causes": issue["probable_causes"], "recommendation": issue["probable_causes"][0]["fix"], "estimated_cost": issue["probable_causes"][0]["cost_range"], } return { "issue": "Unknown", "recommendation": "Schedule on-site diagnostic visit", "estimated_cost": "$95 diagnostic fee", } ``` ## Seasonal Planning Pool services have distinct seasonal phases. The agent manages transitions and prepares for each season. ```python class SeasonalPlanner: SEASONAL_TASKS = { "spring_opening": [ {"task": "Remove cover and clean", "order": 1, "time_min": 30}, {"task": "Inspect equipment (pump, filter, heater)", "order": 2, "time_min": 20}, {"task": "Fill to operating level", "order": 3, "time_min": 15}, {"task": "Prime and start pump", "order": 4, "time_min": 10}, {"task": "Initial chemical treatment (shock)", "order": 5, "time_min": 15}, {"task": "Install ladders and accessories", "order": 6, "time_min": 15}, ], "fall_closing": [ {"task": "Lower water level below returns", "order": 1, "time_min": 20}, {"task": "Blow out plumbing lines", "order": 2, "time_min": 30}, {"task": "Add winterizing chemicals", "order": 3, "time_min": 10}, {"task": "Install winter plugs", "order": 4, "time_min": 15}, {"task": "Install pool cover", "order": 5, "time_min": 30}, {"task": "Disconnect and store pump/filter", "order": 6, "time_min": 20}, ], } def generate_seasonal_schedule( self, pools: list[dict], season: str, start_date: str, ) -> list[dict]: tasks = self.SEASONAL_TASKS.get(season, []) total_time_per_pool = sum(t["time_min"] for t in tasks) pools_per_day = max(1, int(480 / total_time_per_pool)) # 8-hour day schedule = [] for i, pool in enumerate(pools): day_offset = i // pools_per_day schedule.append({ "customer": pool["customer_name"], "address": pool["address"], "scheduled_day": f"Day {day_offset + 1}", "tasks": [t["task"] for t in tasks], "estimated_time_min": total_time_per_pool, }) return schedule ``` ## FAQ ### How does the agent account for different pool types in chemical calculations? The calculations adjust based on pool type (chlorine, saltwater, biguanide) and surface material (plaster, fiberglass, vinyl). Saltwater pools require different alkalinity targets and do not need external chlorine unless the salt cell is underperforming. The agent stores the pool type in the customer profile and applies the correct formula set automatically. ### Can the agent predict when equipment will fail? Yes, through trend analysis. The agent tracks filter pressure readings, pump amperage, and heater cycle counts over time. When pressure rises steadily between cleanings, it indicates filter media degradation. When pump amperage increases, it signals bearing wear. The agent flags these trends 2-4 weeks before likely failure, allowing proactive replacement during scheduled visits. ### How does route optimization handle pools with different service frequencies? Some pools are serviced weekly, others bi-weekly. The agent builds separate route sets for each frequency tier. On weeks when bi-weekly pools are due, it merges them into the weekly route using geographic clustering. This prevents the technician from driving past a bi-weekly pool on the way to a weekly one without stopping. --- #PoolService #ChemicalBalance #ServiceRoutes #EquipmentDiagnostics #SeasonalPlanning #AgenticAI #LearnAI #AIEngineering --- Source: https://callsphere.ai/blog/building-pool-service-agent-maintenance-chemical-balance-repair