Create 4D construction simulations by linking BIM elements to project schedules. Generate time-based visualizations, sequence analysis, and construction phasing with Gantt integration.
中文简介正在准备中。当前显示原文说明。
Manage construction equipment fleet. Track utilization, maintenance, and assignments.
中文简介正在准备中。当前显示原文说明。
实测评分 未实测
查看安装方法安装前可以查看交给 AI 代理的指令内容。
import pandas as pd
from datetime import date, timedelta
from typing import Dict, Any, List, Optional
from dataclasses import dataclass, field
from enum import Enum
class EquipmentStatus(Enum):
AVAILABLE = "available"
IN_USE = "in_use"
MAINTENANCE = "maintenance"
REPAIR = "repair"
RETIRED = "retired"
class EquipmentType(Enum):
CRANE = "crane"
EXCAVATOR = "excavator"
LOADER = "loader"
FORKLIFT = "forklift"
GENERATOR = "generator"
COMPRESSOR = "compressor"
SCAFFOLDING = "scaffolding"
OTHER = "other"
@dataclass
class MaintenanceRecord:
record_id: str
equipment_id: str
maintenance_type: str
scheduled_date: date
completed_date: Optional[date]
cost: float
notes: str = ""
@dataclass
class Assignment:
assignment_id: str
equipment_id: str
project: str
location: str
start_date: date
end_date: Optional[date]
operator: str = ""
@dataclass
class Equipment:
equipment_id: str
name: str
equipment_type: EquipmentType
make: str
model: str
year: int
status: EquipmentStatus
hourly_rate: float
daily_rate: float
current_hours: float = 0
last_maintenance: Optional[date] = None
next_maintenance_hours: float = 500
assignments: List[Assignment] = field(default_factory=list)
class EquipmentFleetManager:
def __init__(self, company_name: str):
self.company_name = company_name
self.equipment: Dict[str, Equipment] = {}
self.maintenance_records: List[MaintenanceRecord] = {}
self._equip_counter = 0
self._assign_counter = 0
def add_equipment(self, name: str, equipment_type: EquipmentType,
make: str, model: str, year: int,
hourly_rate: float, daily_rate: float) -> Equipment:
self._equip_counter += 1
equip_id = f"EQ-{self._equip_counter:04d}"
equip = Equipment(
equipment_id=equip_id,
name=name,
equipment_type=equipment_type,
make=make,
model=model,
year=year,
status=EquipmentStatus.AVAILABLE,
hourly_rate=hourly_rate,
daily_rate=daily_rate
)
self.equipment[equip_id] = equip
return equip
def assign_equipment(self, equip_id: str, project: str, location: str,
start_date: date, operator: str = "") -> Assignment:
if equip_id not in self.equipment:
return None
self._assign_counter += 1
assign_id = f"ASN-{self._assign_counter:04d}"
assignment = Assignment(
assignment_id=assign_id,
equipment_id=equip_id,
project=project,
location=location,
start_date=start_date,
end_date=None,
operator=operator
)
self.equipment[equip_id].assignments.append(assignment)
self.equipment[equip_id].status = EquipmentStatus.IN_USE
return assignment
def return_equipment(self, equip_id: str, hours_used: float):
if equip_id in self.equipment:
equip = self.equipment[equip_id]
equip.status = EquipmentStatus.AVAILABLE
equip.current_hours += hours_used
if equip.assignments:
equip.assignments[-1].end_date = date.today()
def schedule_maintenance(self, equip_id: str, maintenance_type: str,
scheduled_date: date, cost: float):
if equip_id not in self.equipment:
return
record_id = f"MNT-{len(self.maintenance_records) + 1:04d}"
record = MaintenanceRecord(record_id, equip_id, maintenance_type,
scheduled_date, None, cost)
self.maintenance_records[record_id] = record
def get_available_equipment(self, equipment_type: EquipmentType = None) -> List[Equipment]:
available = [e for e in self.equipment.values()
if e.status == EquipmentStatus.AVAILABLE]
if equipment_type:
available = [e for e in available if e.equipment_type == equipment_type]
return available
def get_utilization_report(self) -> Dict[str, Any]:
in_use = sum(1 for e in self.equipment.values()
if e.status == EquipmentStatus.IN_USE)
total = len(self.equipment)
return {
'total_equipment': total,
'in_use': in_use,
'available': sum(1 for e in self.equipment.values()
if e.status == EquipmentStatus.AVAILABLE),
'maintenance': sum(1 for e in self.equipment.values()
if e.status == EquipmentStatus.MAINTENANCE),
'utilization_rate': round(in_use / total * 100, 1) if total > 0 else 0
}
def export_fleet(self, output_path: str):
data = [{
'ID': e.equipment_id,
'Name': e.name,
'Type': e.equipment_type.value,
'Make/Model': f"{e.make} {e.model}",
'Year': e.year,
'Status': e.status.value,
'Hours': e.current_hours,
'Daily Rate': e.daily_rate
} for e in self.equipment.values()]
pd.DataFrame(data).to_excel(output_path, index=False)
fleet = EquipmentFleetManager("ABC Construction")
crane = fleet.add_equipment("Tower Crane #1", EquipmentType.CRANE,
"Liebherr", "280 EC-H", 2020, 150, 1200)
assignment = fleet.assign_equipment(crane.equipment_id, "Office Tower",
"Site A", date.today(), "John Smith")
report = fleet.get_utilization_report()
print(f"Utilization: {report['utilization_rate']}%")
还没有评价。欢迎分享你的使用感受。
无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
Create 4D construction simulations by linking BIM elements to project schedules. Generate time-based visualizations, sequence analysis, and construction phasing with Gantt integration.
中文简介正在准备中。当前显示原文说明。
Analyze 5000+ IFC and Revit projects at scale for patterns, benchmarks, and insights. Big data analysis for construction.
中文简介正在准备中。当前显示原文说明。
Check construction AI systems against the EU AI Act (Regulation 2024/1689): classify risk of estimation, scheduling, CV and agent tools, document transparency, keep human oversight. Use when deploying or auditing AI in construction.
中文简介正在准备中。当前显示原文说明。
Orchestrate multiple AI agents for construction workflows: estimator, scheduler, document, QA and safety agents coordinated by a supervisor agent, with human checkpoints. Use when automating end-to-end project processes with agentic AI.
中文简介正在准备中。当前显示原文说明。
Create Apache Airflow DAGs for construction data pipelines. Orchestrate ETL, validation, and reporting workflows.
中文简介正在准备中。当前显示原文说明。
Automate as-built documentation and digital handover for construction. Compile project records, generate O&M manuals, create asset databases, and ensure complete project closeout.
中文简介正在准备中。当前显示原文说明。