Tiling Trees MCP Server

# Tiling Trees Method - Detailed Examples This document provides complete examples of using the tiling trees method for systematic solution space exploration. ## Example 1: Reducing Food Waste **Problem**: How can we reduce food waste by 50% in urban areas? ### Step 1: Create the tree ``` Tool: create_tree { "name": "Urban Food Waste Reduction", "problemStatement": "How can we reduce food waste by 50% in urban areas within 5 years?" } ``` Returns: Tree with root tile representing "All possible solutions" ### Step 2: First split - by stage in food lifecycle ``` Tool: split_tile { "tileId": "<root-id>", "splitAttribute": "Stage in food lifecycle", "splitRationale": "Food waste occurs at distinct stages with different causes and interventions", "subsets": [ { "title": "Production/Harvesting", "description": "Waste at farms before food enters supply chain - unharvested crops, sorting waste", "isLeaf": false }, { "title": "Processing & Manufacturing", "description": "Waste during food processing, packaging, and manufacturing facilities", "isLeaf": false }, { "title": "Distribution & Retail", "description": "Waste in transportation, warehouses, grocery stores, restaurants before consumer purchase", "isLeaf": false }, { "title": "Consumer Use", "description": "Waste in homes, offices, schools after consumer acquires food", "isLeaf": false } ] } ``` ### Step 3: Validate MECE ``` Tool: mark_mece { "tileId": "<root-id>", "isMECE": true, "coverageNotes": "Covers complete food lifecycle from farm to table. Each stage is distinct with minimal overlap. Edge case: restaurant waste split between 'retail' (prep waste) and 'consumer' (plate waste) - classified as 'distribution' since restaurant is vendor." } ``` ### Step 4: Split "Distribution & Retail" by intervention mechanism ``` Tool: split_tile { "tileId": "<distribution-retail-id>", "splitAttribute": "Intervention mechanism", "splitRationale": "Different mechanisms address different root causes of retail waste", "subsets": [ { "title": "Demand prediction improvement", "description": "Better forecasting to match supply with demand - AI, data analytics", "isLeaf": true }, { "title": "Shelf-life extension", "description": "Technologies to keep food fresh longer - packaging, refrigeration, preservatives", "isLeaf": true }, { "title": "Dynamic pricing", "description": "Price reductions on near-expiry items to increase sales velocity", "isLeaf": true }, { "title": "Donation coordination", "description": "Systems to redirect unsold food to food banks before spoilage", "isLeaf": true } ] } ``` ### Step 5: Evaluate leaf tiles ``` Tool: evaluate_tile { "tileId": "<demand-prediction-id>", "impact": 8, "feasibility": 7, "uniqueness": 6, "timeframe": "1-2 years", "notes": "Many retailers already have data infrastructure for rapid deployment", "calculationsOrPilots": "Pilot with grocery chain reduced waste by 30% using ML demand forecasting. Scales linearly with adoption." } ``` ``` Tool: evaluate_tile { "tileId": "<dynamic-pricing-id>", "impact": 6, "feasibility": 9, "uniqueness": 3, "timeframe": "6 months", "notes": "Low-tech solution, already proven in multiple markets", "calculationsOrPilots": "European grocers using automated markdowns report 15-20% waste reduction. Simple to implement." } ``` ### Step 6: Check coverage ``` Tool: get_coverage_analysis { "treeId": "<tree-id>" } ``` Returns: "3 unexplored tiles - consider splitting these to complete coverage" ### Step 7: Continue with unexplored branches Split "Consumer Use" by intervention type, "Production/Harvesting" by technology, etc. ### Step 8: Find top solutions ``` Tool: get_top_leaves { "criteria": "combined", "limit": 5, "treeId": "<tree-id>" } ``` Returns ranked list of most promising interventions. --- ## Example 2: Improving Solar Panel Efficiency **Problem**: How can we increase solar panel efficiency from 20% to 35%? ### Step 1: Create tree ``` Tool: create_tree { "name": "Solar Panel Efficiency Improvements", "problemStatement": "Increase solar panel energy conversion efficiency from 20% to 35%" } ``` ### Step 2: Split by loss mechanism ``` Tool: split_tile { "tileId": "<root-id>", "splitAttribute": "Energy loss mechanism", "splitRationale": "Physics dictates specific loss mechanisms; addressing each independently", "subsets": [ { "title": "Thermalization losses", "description": "Energy lost when high-energy photons create electron-hole pairs with excess energy dissipated as heat", "isLeaf": false }, { "title": "Sub-bandgap transmission", "description": "Photons with energy below bandgap pass through without generating carriers", "isLeaf": false }, { "title": "Recombination losses", "description": "Electron-hole pairs recombine before collection - radiative, Auger, surface, defect recombination", "isLeaf": false }, { "title": "Optical losses", "description": "Reflection, shadowing from contacts, incomplete absorption", "isLeaf": false }, { "title": "Resistive losses", "description": "Electrical resistance in contacts, semiconductors, interfaces", "isLeaf": false } ] } ``` ### Step 3: Validate (physics-based splits are naturally MECE) ``` Tool: mark_mece { "tileId": "<root-id>", "isMECE": true, "coverageNotes": "Categories based on fundamental energy loss mechanisms in photovoltaics. Comprehensive per Shockley-Queisser analysis. No overlap as each mechanism is physically distinct." } ``` ### Step 4: Split "Thermalization losses" by approach ``` Tool: split_tile { "tileId": "<thermalization-id>", "splitAttribute": "Approach to capture excess energy", "splitRationale": "Different physical mechanisms for harvesting energy that would otherwise thermalize", "subsets": [ { "title": "Multi-junction cells", "description": "Stack multiple cells with different bandgaps to match photon energies more precisely", "isLeaf": true }, { "title": "Hot carrier cells", "description": "Extract carriers before thermalization via energy-selective contacts", "isLeaf": true }, { "title": "Multiple exciton generation", "description": "High-energy photons generate multiple electron-hole pairs (quantum dots, nanocrystals)", "isLeaf": true }, { "title": "Up/down conversion", "description": "Convert photon energies before absorption - combine low-energy photons or split high-energy ones", "isLeaf": true } ] } ``` ### Step 5: Evaluate ``` Tool: evaluate_tile { "tileId": "<multi-junction-id>", "impact": 9, "feasibility": 8, "uniqueness": 4, "timeframe": "3-5 years", "notes": "Proven technology (space applications), needs cost reduction for terrestrial use", "calculationsOrPilots": "Lab cells achieve 47.1% efficiency (NREL). Cost modeling shows path to $0.10/W with manufacturing scale." } ``` ``` Tool: evaluate_tile { "tileId": "<hot-carrier-id>", "impact": 10, "feasibility": 3, "uniqueness": 9, "timeframe": "10+ years", "notes": "Theoretical efficiency >65% but major materials science challenges remain", "calculationsOrPilots": "Proof-of-concept devices show reduced thermalization but not yet practical carrier extraction. Active research area." } ``` --- ## Example 3: Reducing Traffic Congestion **Problem**: How can we reduce urban traffic congestion by 40%? ### Step 1: Create tree ``` Tool: create_tree { "name": "Urban Traffic Congestion Reduction", "problemStatement": "Reduce traffic congestion in major cities by 40% during peak hours" } ``` ### Step 2: Split by intervention category ``` Tool: split_tile { "tileId": "<root-id>", "splitAttribute": "Type of intervention", "splitRationale": "Fundamental approaches to congestion: reduce demand, increase supply, or optimize usage", "subsets": [ { "title": "Reduce travel demand", "description": "Decrease number of trips or miles traveled - telecommuting, land use changes, trip elimination", "isLeaf": false }, { "title": "Shift to other modes", "description": "Move travelers from cars to transit, bikes, walking - modal shift", "isLeaf": false }, { "title": "Increase road capacity", "description": "Add lane-miles of roadway - new roads, wider roads, elevated highways", "isLeaf": false }, { "title": "Optimize existing capacity", "description": "Use current roads more efficiently - traffic management, congestion pricing, smart signals", "isLeaf": false }, { "title": "Shift timing", "description": "Spread demand across time - flexible work hours, off-peak incentives", "isLeaf": false } ] } ``` ### Step 3: Mark MECE ``` Tool: mark_mece { "tileId": "<root-id>", "isMECE": true, "coverageNotes": "Exhaustive list of congestion strategies per transportation economics framework. Categories are mutually exclusive: each solution fits exactly one category based on its primary mechanism." } ``` ### Step 4: Split "Optimize existing capacity" by technology ``` Tool: split_tile { "tileId": "<optimize-capacity-id>", "splitAttribute": "Primary technology/mechanism", "splitRationale": "Different technological approaches to optimization", "subsets": [ { "title": "Adaptive traffic signals", "description": "AI-optimized signal timing based on real-time traffic flow", "isLeaf": true }, { "title": "Congestion pricing", "description": "Variable tolls that increase with traffic density to manage demand", "isLeaf": true }, { "title": "Autonomous vehicle platooning", "description": "Self-driving cars traveling in tight groups to increase effective capacity", "isLeaf": true }, { "title": "Ramp metering", "description": "Controlled highway entrance to maintain optimal mainline flow", "isLeaf": true }, { "title": "Dynamic lane assignment", "description": "Reversible lanes, shoulder use during peak periods", "isLeaf": true } ] } ``` ### Step 5: Realize missing category and add it ``` Tool: add_tiles_to_split { "parentId": "<optimize-capacity-id>", "newTiles": [ { "title": "Route optimization apps", "description": "Navigation systems that distribute traffic across network based on real-time conditions", "isLeaf": true } ] } ``` Note: This invalidates MECE status, need to re-validate. ``` Tool: mark_mece { "tileId": "<optimize-capacity-id>", "isMECE": true, "coverageNotes": "Now complete with route optimization included. Each approach uses distinct mechanism for optimization." } ``` --- ## Example 4: Discovering a Missing Category Mid-Exploration This example shows how the method helps identify gaps in thinking. **Problem**: Improve smartphone battery life by 3x ### Initial split - by approach type ``` Tool: split_tile { "tileId": "<root-id>", "splitAttribute": "Approach type", "splitRationale": "Fundamental categories of battery improvement", "subsets": [ { "title": "Increase battery capacity", "description": "More mAh in same volume - better chemistry, materials", "isLeaf": false }, { "title": "Reduce power consumption", "description": "Make components more efficient - CPU, display, radio", "isLeaf": false } ] } ``` ### Try to validate MECE ``` Tool: mark_mece { "tileId": "<root-id>", "isMECE": false, "coverageNotes": "INCOMPLETE - missing category for power delivery efficiency (voltage regulation losses, charging efficiency). Need to add." } ``` ### Add missing category ``` Tool: add_tiles_to_split { "parentId": "<root-id>", "newTiles": [ { "title": "Improve power delivery efficiency", "description": "Reduce losses in voltage regulation, charging circuits, power management", "isLeaf": false } ] } ``` ### Now validate ``` Tool: mark_mece { "tileId": "<root-id>", "isMECE": true, "coverageNotes": "Now complete: (1) more energy stored, (2) less energy consumed, (3) less energy lost in delivery. These three categories are exhaustive and mutually exclusive." } ``` **Key insight**: The MECE validation process forced us to think systematically and identify a category we initially overlooked! --- ## Example 5: Complex Multi-Level Exploration **Problem**: Reduce healthcare costs by 30% This demonstrates a deeper tree with multiple split levels. ### Level 1: Split by cost category ``` Tool: split_tile { "tileId": "<root-id>", "splitAttribute": "Cost category", "splitRationale": "Healthcare spending分为distinct economic categories", "subsets": [ {"title": "Prevention/Wellness", "description": "Costs before disease onset"}, {"title": "Diagnosis", "description": "Identifying health conditions"}, {"title": "Treatment", "description": "Curing or managing conditions"}, {"title": "Administration", "description": "Insurance, billing, overhead"} ] } ``` ### Level 2: Split "Treatment" by intervention type ``` Tool: split_tile { "tileId": "<treatment-id>", "splitAttribute": "Intervention type", "splitRationale": "Different medical intervention modalities", "subsets": [ {"title": "Pharmaceuticals", "description": "Drug-based treatments"}, {"title": "Surgery", "description": "Invasive procedures"}, {"title": "Therapy", "description": "Physical, occupational, psychological therapy"}, {"title": "Devices", "description": "Medical devices (pacemakers, prosthetics, etc.)"} ] } ``` ### Level 3: Split "Pharmaceuticals" by cost reduction mechanism ``` Tool: split_tile { "tileId": "<pharmaceuticals-id>", "splitAttribute": "Cost reduction mechanism", "splitRationale": "Distinct approaches to reducing drug costs", "subsets": [ {"title": "Generic adoption", "description": "Increase use of off-patent generics", "isLeaf": true}, {"title": "Negotiation/regulation", "description": "Price controls, bulk purchasing", "isLeaf": true}, {"title": "Manufacturing efficiency", "description": "Cheaper production processes", "isLeaf": true}, {"title": "Drug efficacy improvement", "description": "Better drugs need shorter treatment", "isLeaf": true} ] } ``` Now you can evaluate each leaf and work through other branches similarly. --- ## Example 6: Revisiting a Tree as Context Changes **Scenario**: You created a tiling tree for "reduce building energy consumption" in 2020. It's now 2025 and heat pump technology has advanced significantly. ### Original evaluation (2020) ``` Tool: evaluate_tile { "tileId": "<heat-pump-id>", "impact": 7, "feasibility": 5, "uniqueness": 6, "timeframe": "5-10 years", "notes": "Promising but high upfront cost limits adoption", "calculationsOrPilots": "ROI analysis shows 12-year payback period" } ``` ### Updated evaluation (2025) ``` Tool: evaluate_tile { "tileId": "<heat-pump-id>", "impact": 8, "feasibility": 9, "uniqueness": 5, "timeframe": "1-2 years", "notes": "Technology matured, costs dropped 60%, policy incentives available", "calculationsOrPilots": "New ROI analysis shows 3-year payback with subsidies. Multiple successful deployments." } ``` **Key lesson**: Revisit trees periodically. Technologies and contexts evolve, turning previously unviable branches into top solutions! --- ## Example 7: Using Validation to Catch and Fix Antipatterns **Scenario**: You're building a tree for renewable energy solutions and make some common mistakes. The validation tools catch them. ### Initial problematic split ``` Tool: split_tile { "tileId": "<root-id>", "splitAttribute": "Energy type", "splitRationale": "Different energy sources have different characteristics", "subsets": [ { "title": "Solar power", "description": "Using sunlight to generate electricity", "isLeaf": false }, { "title": "Wind power", "description": "Natural wind turning turbines", "isLeaf": false }, { "title": "Advanced renewable technologies", "description": "Newer, innovative approaches to renewable energy", "isLeaf": false }, { "title": "Other renewables", "description": "Miscellaneous other renewable energy sources", "isLeaf": false } ] } ``` ### Run validation ``` Tool: validate_split_quality { "tileId": "<root-id>" } ``` ### Response reveals multiple issues: ```json { "score": 40, "issues": [ { "type": "vague_language", "severity": "warning", "message": "Tile 'Wind power' uses vague term 'natural' which may lack precision", "tileId": "<wind-tile-id>", "suggestion": "Replace 'natural' with measurable properties..." }, { "type": "vague_language", "severity": "warning", "message": "Tile 'Advanced renewable technologies' uses vague term 'advanced'", "tileId": "<advanced-tile-id>", "suggestion": "Replace 'advanced' with measurable properties..." }, { "type": "catch_all_bucket", "severity": "error", "message": "Tile 'Other renewables' appears to be a catch-all bucket", "tileId": "<other-tile-id>", "suggestion": "Replace with specific, well-defined categories..." } ], "recommendations": [ "Address 1 critical issue(s) before proceeding", "Review 2 warning(s) to improve split quality", "Replace vague terms with measurable physical properties", "Replace catch-all categories with specific, splittable subsets" ] } ``` ### Fix the issues - revised split ``` Tool: split_tile { "tileId": "<root-id>", "splitAttribute": "Primary energy conversion mechanism", "splitRationale": "Categorizing by the fundamental physical process that converts energy", "subsets": [ { "title": "Photovoltaic conversion", "description": "Direct conversion of photons to electricity via semiconductor effect", "isLeaf": false }, { "title": "Thermal conversion", "description": "Heat-based energy conversion (concentrated solar thermal, geothermal)", "isLeaf": false }, { "title": "Kinetic energy harvesting", "description": "Converting motion to electricity - wind, hydroelectric, tidal, wave", "isLeaf": false }, { "title": "Chemical/biochemical conversion", "description": "Energy from chemical reactions - biomass, biogas, algae", "isLeaf": false }, { "title": "Gravitational potential", "description": "Using elevation differences - pumped hydro storage", "isLeaf": false } ] } ``` ### Validate again ``` Tool: validate_split_quality { "tileId": "<root-id>" } ``` ### Much better result: ```json { "score": 100, "issues": [], "recommendations": [ "Split appears well-structured" ] } ``` ### Key improvements made: 1. **Removed vague language**: "Natural" → "motion-based"; "Advanced" → specific physical processes 2. **Eliminated catch-all bucket**: "Other renewables" → Complete MECE categories based on physics 3. **Used physics-based split**: Changed from solution types to energy conversion mechanisms 4. **Precise definitions**: Each category has clear physical meaning 5. **Complete coverage**: All renewable energy fits exactly one category ### Getting tree-wide validation ``` Tool: get_tree_validation_report { "treeId": "<tree-id>" } ``` Returns overall score and all split reports, helping identify weak spots across the entire tree. --- ## Tips from These Examples 1. **Start with clear problem statements** - Specific, measurable goals 2. **Use physics/economics/math-based splits** when possible - Natural MECE categories 3. **Validate splits immediately** - Use `validate_split_quality` after every split to catch antipatterns early 4. **Avoid vague language** - Replace "natural", "advanced", "traditional" with measurable properties 5. **Never use catch-all buckets** - "Other" categories prevent systematic exploration 6. **Document edge cases** - Record how boundary cases are classified 7. **Don't rush to leaves** - Explore breadth before depth 8. **Evaluate with data** - Use calculations and pilots, not just intuition 9. **Check tree-wide quality** - Use `get_tree_validation_report` to find weak spots 10. **Revisit periodically** - Context changes over time