Welcome to DynMEP! I'm Alfonso Davila, an electrical engineer and automation enthusiast with 20+ years of experience in power systems and multidisciplinary project leadership. My mission is to revolutionize MEP and BIM workflows through innovative automation tools, computational methods, and renewable energy solutions. From Revit MEP scripts to AI-powered symbol detection and advanced mathematical research, these projects save time, enhance accuracy, and drive sustainability. All repositories are hosted at github.com/DynMEP under the MIT License.

With 20+ years in electrical engineering specializing in power distribution, MEP design, and sustainable energy systems, I develop cutting-edge tools across multiple domains:

• Revit MEP Automation: NEC-compliant Dynamo scripts saving 5–15 hours per project
• Computer Vision: YOLO-based AI for automated technical drawing analysis
• Renewable Energy: MATLAB simulations for hybrid solar/wind integration
• Motor Control Systems: VFD simulation and analysis for industrial applications
• Reinforcement Learning: Deep Q-Networks for microgrid energy optimization
• Computational Mathematics: Advanced algorithms for additive combinatorics research
• Professional Focus: Power systems, BIM workflows, and green energy solutions

Repository: github.com/DynMEP/DynMEP

A comprehensive collection of Python/Dynamo scripts designed to turbocharge Revit MEP workflows with NEC 2023 compliance.

Available Scripts:

File: Clash_Detection_Electrical_HVAC_Dynamo.py
Automates clash detection between electrical systems (conduits, cable trays) and HVAC ducts in Revit 2025. Generates detailed CSV reports, identifies conflicts early, and reduces costly rework.

• Time Savings: 10–15 hours per project
• Compliance: NEC-compliant design verification
• Output: CSV reports with detailed clash locations

File: Automate_Lighting_Fixture_Placement.py
Places fixtures in rooms using grid and boundary logic with DIALux CSV integration. Ensures ceiling-hosted installations per NEC 410.36(A).

• Time Savings: 6–10 hours per project
• Features: Grid-based placement, wall conflict avoidance
• Output: CSV logs for compliance verification

File: Receptacle_Placement_Revit.py
Automates NEC-compliant receptacle placement on both wall faces with precise spacing calculations.

• Time Savings: 6–10 hours per project
• Compliance: Meets NEC spacing standards
• Output: Detailed logs and CSV documentation

File: Panel_Schedule_Generator_Dynamo.py
Generates NEC-compliant panel schedules with automated circuit sorting and voltage correction.

• Time Savings: 10–15 hours per project
• Features: Load and current data validation
• Output: Detailed CSV schedules

File: Electrical_Load_Estimator_Dynamo.py
Automates load estimation for electrical, HVAC, and lighting systems with NEC compliance.

• Time Savings: 5–8 hours per project
• Features: Comprehensive error handling
• Output: Formatted tables and CSV reports

File: Conduit_Length_Calculator_Dynamo.py
Calculates conduit lengths for material takeoffs and cost estimation.

• Time Savings: 8–12 hours per project
• Features: Type and size summaries
• Output: Detailed CSV reports

Repository: github.com/DynMEP/YOLOplan

YOLOplan revolutionizes technical drawing analysis by automating symbol detection and counting using advanced YOLO object detection. It processes PDF, image, and CAD formats with exceptional accuracy, even in noisy or complex plans.

Key Features:

• Automatic Detection: Identifies and counts electrical, HVAC, plumbing, and architectural symbols
• Multi-Format Support: Works with PDF, images, and CAD files
• Robust Performance: Handles background noise, varied symbol sizes, and rotations
• Custom Training: Train models for your own symbols
• Export Options: CSV, Excel, JSON, or annotated images with bounding boxes

Applications:

• Electrical and MEP engineering takeoff
• Construction estimation
• BIM and digital building modeling
• Facility management
• QA/QC for technical plans

Technology Stack:

• Python 3.8+
• Ultralytics YOLO
• OpenCV, pdf2image, PyMuPDF

Project Structure:

YOLOplan/
├── yolo_plan_core/        # Common utilities
├── yolo_plan_electric/    # Electrical symbols models
├── yolo_plan_hvac/        # HVAC models (future)
├── datasets/              # Training datasets
├── notebooks/             # Demo notebooks
└── demo_takeoff_electric.ipynb

Repository: github.com/DynMEP/GreenPowerHub

GreenPowerHub serves as a comprehensive hub for renewable energy integration tools, simulations, and resources. Focused on advancing sustainable energy solutions through power systems expertise and advanced simulations.

Mission:

• Develop innovative tools for renewable energy integration
• Enhance power system efficiency through data-driven simulations
• Foster global collaboration for sustainable practices
• Drive positive impact with cutting-edge energy solutions

Current Scripts:

File: Hybrid_Renewable_Integration.m
Released: October 07, 2025

A MATLAB script simulating hybrid solar PV and wind energy systems integrated into power distribution grids.

Features:

• Calculates power outputs based on irradiance, temperature, and wind speed
• Applies basic MPPT (Maximum Power Point Tracking) approximation
• Estimates daily energy yield and grid contribution
• Saves results to CSV for MEP integration

Benefits:

• Time Savings: 5–10 hours per analysis
• Compliance: Supports NEC 2023-compliant designs
• Planning: Aids renewable project feasibility studies

Planned Development:

• Renewable_Load_Optimizer.py: Python script for load distribution optimization with Dynamo integration
• NEC 220 demand factor calculations
• Enhanced grid reliability analysis

Requirements:

• MATLAB (base version, no toolboxes required)
• Python 3.x (for future scripts)
• Optional: Revit/Dynamo for MEP integration

Repository: github.com/DynMEP/vfd-motor-simulation

A comprehensive Python-based simulation of Variable Frequency Drive (VFD) controlled motor startup for high-power induction motors, with comparative analysis against Direct-On-Line (DOL) starting methods.

Target Audience:

• Electrical Engineers (system design and analysis)
• Research Applications (motor control studies)
• Educational Purposes (understanding VFD operation)
• Industrial Planning (equipment specification)

Key Features:

• Accurate Physics Modeling: Proper torque-slip characteristics with constant V/f control
• Multiple Load Types: Constant torque, fan/pump (quadratic), and constant power loads
• VFD vs DOL Comparison: Side-by-side performance analysis with 81% peak current reduction
• Real-time Analysis: Speed, torque, current, slip, power, and efficiency tracking
• Data Export: Timestamped CSV files for further analysis
• Professional Visualizations: 6-panel comparison dashboard

Simulation Capabilities:

• 800HP (596.6 kW) three-phase induction motor (configurable)
• Multiple load type simulations (conveyors, fans/pumps, machine tools)
• DOL starting comparison with grid impact assessment
• Energy and efficiency analysis during startup
• Comprehensive performance metrics

Key Results (800HP motor example):

VFD Starting:
- Peak Current: 1104.6 A (1.23 × FLA)
- Final Speed: 1729 RPM
- Final Slip: 3.93%
- Startup Energy: 4.424 kWh

DOL Starting:
- Peak Current: 5821.6 A (6.50 × FLA)
- Starting Time: ~5 seconds
- Startup Energy: 2.724 kWh

VFD Advantages:
- 81% lower peak current
- Significantly reduced mechanical stress
- Minimal voltage sag (<5% vs 30-40%)

Load Type Options:

1. Constant Torque: Conveyors, hoists, positive displacement pumps
2. Fan/Pump: Centrifugal fans and pumps (torque ∝ speed²)
3. Constant Power: Machine tools, winders (torque ∝ 1/speed)

Technical Specifications:

• Python 3.x with NumPy, SciPy, Matplotlib
• First-order induction motor model with torque-slip characteristics
• Constant V/f control with low-frequency voltage boost
• IEEE standards validated
• Execution time: ~0.6 seconds for standard simulation

Applications:

• VFD rating determination
• Electrical infrastructure capacity evaluation
• Mechanical system compatibility assessment
• Energy consumption comparison
• Educational demonstrations
• Research baseline for advanced control strategies

Version: 3.0.0 (October 10, 2025)

Future Roadmap:

• GUI interface with real-time parameter adjustment (Q1 2026)
• Additional motor types and harmonic analysis (Q1-Q2 2026)
• Multi-motor simulation and thermal modeling (Q2 2026)
• Real-time hardware integration and web dashboard (Future)

Repository: github.com/DynMEP/dynamic-microgrid-resilience

An anticipatory Deep Q-Network (DQN) approach that achieves near-perfect microgrid energy management through intelligent battery control. The system learns to prepare for evening peak demand hours in advance, achieving 100% load coverage with zero unmet energy.

Target Audience:

• Energy Engineers (microgrid design and optimization)
• Renewable Energy Researchers (AI-based control systems)
• Utility Companies (grid reliability and peak management)
• Educational Institutions (reinforcement learning applications)

Key Features:

• 98.8% Average Load Coverage: Across 100 diverse test scenarios
• 100% Best Policy Performance: Optimal policy found at episode 100
• Anticipatory Behavior: Pre-evening preparation with 87% SOC by hour 15
• Time-to-Event State Augmentation: Explicit temporal encoding for peak anticipation
• Hierarchical Reward Shaping: Extreme penalties for evening failures, bonuses for preparation
• Fast Training: 21 minutes on laptop CPU to achieve optimal performance

System Configuration:

• Solar PV Array: 5 kW DC capacity
• Wind Turbine: 3 kW AC capacity
• Battery Storage: 18 kWh capacity, 6 kW power rating
• NEC 2023 Compliant (Articles 690, 694, 706)
• 95% round-trip efficiency

Performance Highlights:

✅ 98.8% Average Load Coverage (100 test scenarios)
✅ 100% Peak Policy Performance
✅ 99.2% Evening Reliability (vs 71.5% baseline)
✅ 0.13 kWh Average Unmet Energy
✅ 4.3 Year Payback Period
✅ 190% ROI (20-year projection)
✅ $7,088 Annual Savings

Technology Stack:

• Python 3.8+
• PyTorch 2.0+ (Deep Q-Network)
• NumPy, Pandas, Matplotlib
• Experience Replay with Target Networks
• 3-Layer Neural Network (256-128-64)

Quick Start:

git clone https://github.com/DynMEP/dynamic-microgrid-resilience.git
cd dynamic-microgrid-resilience
pip install -r requirements.txt

# Quick demo (1 minute)
python cli.py demo

# Full training with plots
python cli.py train --full --plot --economics

# Evaluate pre-trained model
python cli.py evaluate --auto-latest --scenarios 100 --plot

Applications:

• Off-grid residential/commercial microgrids
• Peak demand management and load shifting
• Renewable energy integration optimization
• Battery energy storage system (BESS) control
• Grid resilience and reliability improvement
• Energy cost reduction strategies

Economic Viability:

• Total System Cost: $35,808
• Payback Period: 4.3 years (vs 7-10 industry standard)
• 20-Year NPV: $57,886 (5% discount rate)
• Internal Rate of Return: 21.4%

Research Paper: "Anticipatory Deep Reinforcement Learning for Microgrid Energy Management: Achieving 100% Evening Peak Coverage Through Pre-Event Preparation" by Alfonso Davila (2025)

Repository: github.com/DynMEP/ZeroSumFreeSets-Z4

An advanced computational mathematics project providing explicit constructions of maximal 3-zero-sum-free subsets in (ℤ/4ℤ)ⁿ. This addresses an open problem in additive combinatorics posed by Nathan Kaplan in the 2014 CANT problem session.

Mathematical Problem: Find the largest subset H ⊆ (ℤ/4ℤ)ⁿ such that there are no distinct x, y, z ∈ H with x + y + z ≡ 0 (mod 4) (pointwise vector addition).

Breakthrough Results:

• Proven Optimal Density: 0.5 for all n
• Universal Construction: First coordinate odd (1 or 3 mod 4)
• Explicit Sizes:
  • n=5: 512 vectors (50% of 1024)
  • n=6: 2048 vectors (50% of 4096)
  • n=7: 8192 vectors (50% of 16384)

Methodology: AI-assisted hybrid greedy-genetic algorithm refined with Grok (xAI), inspired by combinatorial searches like FunSearch.

Algorithm Evolution:

1. Baseline (baseline_script.py): Initial greedy approach (176 for n=5)
2. Refined (refined_script.py): Enhanced with probes (512 for n=5)
3. Omni-Optimized v5.0.0 (omni_optimized_hybrid_discovery_v5.py):
   • Universal optimal construction
   • Stratified sampling and adaptive mutations
   • GPU acceleration with batching (n=8+)
   • Early stopping and profile chaining

Features:

• Exhaustive verification of 3-zero-sum-free property
• JSON export of full subsets with metadata
• GPU support for large-scale computation
• Comprehensive documentation and references

Installation:

git clone https://github.com/DynMEP/ZeroSumFreeSets-Z4.git
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
python3 omni_optimized_hybrid_discovery_v5.py --n 6 7 --mod 4 --profile all

Academic References:

• Y. Caro, A weighted Erdős–Ginzburg–Ziv theorem, J. Combin. Theory Ser. A (1997)
• W.D. Gao and A. Geroldinger, Zero-sum problems survey, Expo. Math. (2006)
• S.J. Miller et al., CANT Problem Sessions (2017)
• MathOverflow discussion thread available

Release: v5.0.0 (August 28, 2025) - Universal construction with GPU support

• Revit Scripts: Autodesk Revit 2025, Dynamo 2.17+, Python (IronPython 2.7 or CPython 3)
• YOLOplan: Python 3.8+, Ultralytics YOLO, OpenCV
• GreenPowerHub: MATLAB (base version) or Python 3.x
• vfd-motor-simulation: Python 3.x, NumPy, SciPy, Matplotlib
• ZeroSumFreeSets: Python 3.8+, optional PyTorch for GPU

git clone https://github.com/DynMEP/DynMEP.git

1. Open Revit 2025 and load your MEP project
2. Launch Dynamo and create a new workspace
3. Add a Python Script node and paste the desired script
4. Connect appropriate inputs (see individual script documentation)
5. Run and check Desktop outputs (CSV, logs)

git clone https://github.com/DynMEP/YOLOplan.git
cd YOLOplan
pip install -r requirements.txt

Run detection on your technical drawings and export results to CSV/Excel.

git clone https://github.com/DynMEP/GreenPowerHub.git

Open Hybrid_Renewable_Integration.m in MATLAB, adjust parameters, and run.

git clone https://github.com/DynMEP/vfd-motor-simulation.git
cd vfd-motor-simulation
pip install numpy scipy matplotlib
python vfd_simulation_v3.py

Customize motor parameters (HP, voltage, poles) and load types in the configuration section.

git clone https://github.com/DynMEP/ZeroSumFreeSets-Z4.git
python3 omni_optimized_hybrid_discovery_v5.py --n 6 --runs 20

Time Savings Across Projects:

• Revit MEP Scripts: 5–15 hours per project
• YOLOplan: Automated symbol counting (hours to minutes)
• GreenPowerHub: 5–10 hours per renewable energy analysis
• vfd-motor-simulation: Instant VFD analysis vs. manual calculations (hours saved)
• Dynamic Microgrid: AI-powered energy management with 4.3-year payback
• ZeroSumFreeSets: Breakthrough mathematical research results

Professional Value:

• NEC Compliance: All electrical scripts meet NEC 2023 standards
• Accuracy: Reduce errors through automation
• Innovation: Cutting-edge AI and mathematical methods
• Sustainability: Support renewable energy adoption
• Open Source: MIT License for maximum collaboration

Industry Impact:

• MEP engineering automation
• BIM coordination excellence
• Renewable energy integration
• Motor control and VFD system design
• Academic research advancement
• Construction estimation efficiency

All repositories are licensed under the MIT License. See individual repository LICENSE files for details.

Contribute: Fork any repository, submit pull requests, or open issues to enhance these tools.

Custom Development: Need specialized automation, AI models, or renewable energy simulations? Contact me for consulting and custom solutions.

Support: Report issues via GitHub Issues on the respective repositories.

Community: Join the growing community of engineers, developers, and researchers using these tools worldwide.

Author: Alfonso Antonio Dávila Vera
Email: davila.alfonso@gmail.com
LinkedIn: www.linkedin.com/in/alfonso-davila-3a121087
GitHub: github.com/DynMEP
YouTube: @DynMEP
Website (Coming Soon): dynmep.com

• 20+ Years: Electrical engineering expertise
• Multiple Domains: MEP automation, AI/ML, renewable energy, mathematics
• Open Source: All tools freely available under MIT License
• Innovation: Combining traditional engineering with cutting-edge technology
• Impact: Serving MEP firms, engineers, and researchers globally
• Validated: IEEE standards compliance and field-tested results

"Let's build smarter MEP workflows, advance renewable energy, and push the boundaries of computational research together! 🚧🌿🔬"

New scripts and repositories are added regularly - check back often!