Advanced Digitalization of Reclaimed Material
INDEX
Context
PART 1
Pre-demolition material dataset processing
Demonstrator development
Commercialization
PART 2
Design demonstrator development
Design strategy from post-demolition dataset
Post-demolition items assessment
This project investigates reusability potentials from post and pre-demolitions sites. It is divided into 2 parts, the first one aims to create an innovative digital decision support for deconstruction and demolition. It employs an automatic digitization method for the near-end life cycle stage of a building and considers it as a source of high value assets. The second part of the research links a
post-demolition database with a computational design tool that matched the designed components with relevant stored materials by their design requirements, as well as providing suggestions for design changes. The proposed iterations aim to optimize repurposed material utilization, performance, and cost. The research reduces the environmental impact of the sector by promoting the use of locally sourced, readily available, and reclaimed materials in an automated way. It exposes the potential for constructing datasets of reclaimed materials, digitally available, for sharing and organizing material harvesting and facilitating its incorporation in new designs.
CONTEXT
RESEARCH DOMAIN
Design
Construction
Operation
Demolition
PROBLEM
PROBLEM
MATERIAL VALUE
330 EUROS Raw Recycled
20 EUROS 5 EUROS
Steel
Concrete
10 EUROS
Wood
10 EUROS
Masonry
Bricks
Service Contract on Construction and Demolition Waste European Demolition Association European Regional Development Fund https://www.kurtz-bros.com/recycled-aggregate/ Eurostat Statistics Explained
OPPORTUNITIES
WASTE REDUCTION
By 2020 the European goal for recovery rate was set to 70%, currently some countries are still on 10%
European commission Waste Framework Directive
MATERIAL FOOTPRINT
BUILDING STOCK AGING
The global material footprint grew from 73.2 billion metric tons in 2010 to 85.9 billion metric tons
Approximately 42% of the buildings in Spain have more than 80 years and there are 10000 demolitions per year just in Spain
in 2017, a 17.4 % increase.
Ensure sustainable consumption and production patterns. SDG Report
Atlas Estadístico de la Construcción y Vivienda en España
19% 42% Pre-1960 Material flow from building stock
1961-1990 1991-2010
39%
CHALLENGES
1
LACK OF INFORMATION ON THE QUANTITY AND LOCATION OF THE MATERIALS
2
LACK OF TRUST IN REUSE QUALITY
3
LACK OF INTEGRATION AMONG STAKEHOLDERS
Construction material waste million tones
Service Contract on Construction and Demolition Waste European Demolition Association European Regional Development Fund Eurostat Statistics Explained
INTERVIEWS
RESEARCH AIMS
A) Material Report
B) Design Strategy
PRE-DEMOLITION DATASET
POST-DEMOLITION DATASET
STAKEHOLDERS
A) Material Report
B) Design Strategy
PRE-DEMOLITION DATASET
POST-DEMOLITION DATASET
Secondary Materials reclaimed product trader-provider
Designers & Planners
Deconstruction Company
Supplier of secondary products and materials
Products Maintenance Services
Material Report
Material Health Score
Location and quantity of the materials
Where the material can be used in the design
T E S A T A D L A I R E T A A|M
In this section the project develops an innovative digital decision support system on 5 demonstrators, integrating various digital tools and processes (3D scanning, imagery localization and point cloud classification) that helps to define the most sustainable and economical deconstruction and reuse strategy for a building. The results showed that it is possible to quantify, qualify and map the flows of products and materials from buildings undergoing renovation, redevelopment or deconstruction and to assess, even before the work begins, their potential for reuse and recycling.
CURRENT SITUATION
USUAL BUSINESS
TRUST
BATIRIM
REGULAR DEMOLITION
MANUAL SELECTIVE DEMOLITION/ DISMANTLING
DIGITIZED SITE INSPECTION WITH AUTOMATIC ANALYSIS (DIGITAL DECONSTRUCTION)
PROPOSITION
DIGITAL SUPPORT SYSTEM FOR DECONSTRUCTION AND DEMOLITION
Increase HQ upcycling
+
Increase the recovery rate and volume from building waste.
Confidence Reliability
+
Improve the quality, value and security of upcycled material supply
Stakeholders Integration
+
Secondary raw materials and reuse components marketplace
MATERIAL DATASET IMPLEMENTATION
1
2
3
4
5
Building Inspection
Geometric Reconstruction
Classification & Localization
Integration
Planning Interface
MATERIAL DATASET IMPLEMENTATION
3
4
2
DEMONSTRATORS INDEX
TIMELINE
TECHNOLOGY DEVELOPMENT
ATELIER INTERIOR
MAIN BUILDING ROOF STRUCTURE
IAAC ROOF PAVILION
SCALED EXAMPLE SITE
BCN DEMO SITE
TECHNOLOGY DEVELOPMENT
TIMELINE
TECHNOLOGY READINESS
ATELIER INTERIOR
MAIN BUILDING ROOF STRUCTURE
IAAC ROOF PAVILION
SCALED EXAMPLE SITE
BCN DEMO SITE
SCALABILITY
DEMONSTRATOR DEVELOPMENT
BUILDING INSPECTION BARCELONA DEMO SITE : CREU DE MOLERS 10
Built 1900 196 square meters Fitout removed
STANDARD
DEVIATIONS
Brick walls and columns Concrete floors Wooden beams
Painted brick walls Corrugated ceiling that resembles metal
WORKFLOW
TECHNOLOGY COMPARISON
TECHNOLOGY COMPARISON
TECHNOLOGY COMPARISON
TECHNOLOGY COMPARISON
PHOTOGRAMMETRY
Texture Quality
Consistency Column Profile
Planarity
MOBILE LIDAR
STATIONARY LIDAR
TECHNOLOGY COMPARISON
Type
Photogrammetry Mobile Lidar Stationary Lidar
Time
Resolution
Texture Quality
Planarity
Consistency
Number of points
High Fidelity Imagery
BUILDING INSPECTION
GEOMETRIC RECONSTRUCTION
Site photography
Photogrammetry point clouds
Volumetric BIM objects
GEOMETRIC RECONSTRUCTION
Raw point cloud
Heuristic association with bim types
Ransac shape detection and cloud splitting
Merging and volumizing to site geometry
PLANE OBJECT SEGMENTATION
Prescaling via local density Truss from ransac
Initial segmentation with dbscan
Simple k-means clustering
Assignment by local direction
Final segmentation with merging and face estimation
BIM OBJECT JOINING
Simple K-means clustering by location produces errors Slabs and walls are represented simply
Pre-transforming coordinates via Galapagos-produced matrix
Beams and columns may contain many planes and must be joined
Known beams end up with recognizable clusters
CLASSIFICATION & LOCALIZATION
CLASSIFICATION & LOCALIZATION
CLASSIFICATION MODEL
CLASSIFICATION MODEL
BOVW
CNN
LOCAL BINARY PATTERNS
HUE AND SATURATION
STACK
PERPLEXITY
5
15
30
50
INTEGRATION LOCALIZING WITHIN POINT CLOUDS
POINT CLOUD
RENDER ALIGNED VIEWS WITH SPARSITY MITIGATION
PERFORM IMAGE LOCALIZATION
PROJECT LOCALIZATION MAPS BACK INTO MESH
INTEGRATION
LOCALIZATION MAP
EXTRINSIC MATRIX
INTRINSIC MATRIX
INTEGRATION
Projects entire material map rather than random sampling Applies to all scene geometry at once
Texture atlas generated for all scene elements
INTEGRATION
Transform map into world space
Localized model from intersections
Site element ID Color UV coordinates
Cast rays from camera through model
Material association
Texture reconstruction
PLANNING INTERFACE
ROUGH DIMENSIONS END USER NEEDS
DIMENSIONS GEOMETRIC ANALYSIS
AESTHETICS
TEXTURED MESH
STRUCTURE
WEBVIEWER INTERFACE
FACE TEXTURES
DEFECT LOCALIZATION MYSQL DATABASE
OTHER INTERFACES
USER INTERFACE DEMO
PLATFORM VISION
TIMELINE
SITE PHOTOGRAPHY
SINGLE IMAGE CLASSIFICATION
PHOTOGRAMMETRY
PER-ELEMENT MATERIAL ASSOCIATIONS
PATCHED LOCALIZATION
MANUAL RECONSTRUCTION
LOCAL SQLITE STORAGE
EXPANDED CLASSIFICATION
SIMPLE PLANE EXTRACTION
AWS BASED MYSQL STORAGE
ELEMENT-BASED BIM RECONSTRUCTION
DEMO WEB-VIEWERS
GEOMETRY-INFORMED CLASSIFICATION
INTEGRATION WITH IFC STANDARDS
PROJECT-ALIGNED WEB VIEWER
B
DESIGN STRATEGY
The aim of this section is to augment the design and construction potentials of irregular materials as well as leverage digital means toward advancing the widespread use of reclaimed materials in structurally predictable assemblies. The method relies on transferring a secondary raw material database to designers and builders to promote the use of valuable secondary-source materials and better inform early design decisions when repurposing construction waste. The initial tested material will be wood in batten shapes, and the design will be guided by a “form follows availability” strategy focusing on sparse and local availability. Therefore a local and small quantity of material will be used and every item will be logged and compared.
STATE OF THE ART
FORM FOLLOWS AVAILABILITY – DESIGNING STRUCTURES THROUGH REUSE
(Brütting, Senatore and Fivet, 2019)
CYCLOPEAN CANNIBALISM
(Clifford and Mcgee, 2019)
DIGITAL FABRICATION OF STANDARDLESS MATERIALS
(MacDonald, Schumann, Hauptman, 2019)
WORKFLOW Feedback from scarcity
Material
User
Negotiation of form between author and non-standard material
Digital Inventory of ready to assemble pieces
Dataset 3D mesh
Generative Design
Element Matching
Digital fabrication of parts
Optimized Design
Assembly of structure
DATASET
85% REUSABLE 15% RECYCLE
XX 10 CHAIRS
X7 BUS STOPS
POST DEMOLITION ITEMS ASSESSMENT
1
Element Inspection with Unsorted Input
2
Geometric and Textural Extraction
3
Geometric and Textural Analysis
4
Integration to Database
+
PRECISION
+
FEATURES
+
CONFIDENCE
+
TRACEABILITY
ELEMENT INSPECTION WITH UNSORTED INPUT ROBOT SETUP
1 | Scanning procedure
2 | Reorienting Piece
SCANNING AUTOMATION SETUP
Scanning Area
ENDTOOL | SECOND ITERATION
Avoid Collision
45°
20° Rotation
PICK AND PLACE
Scan of pickup area with stick arbitrarily placed
As sticks are thin, best-fit lines can orient bounding boxes
Scan with z-level cutoff, leaving stick points
Evaluated plane from bounding box used for pickup target
GEOMETRIC AND TEXTURAL EXTRACTION
FINAL CROPPED CLOUD HALF OF STICK CAPTURED EACH PASS
SCANS REORIENTED INTO WORLD SPACE
RESOLUTION OF COMPLETE SCAN WITH RGB CHANNEL
REALSENSE METHOD QUALITY UPDATE
ELEMENT ID : 45a0d07f-4d69-46bc-a642-c8ca98e471bc LENGTH : 794 mm
ELEMENT INSPECTION WITH UNSORTED INPUT PHOTO SETUP
AUTOMATIC SUBJECT MASKING
ELEMENT INSPECTION WITH UNSORTED INPUT WORKFLOW
.NEF Photos
Exposure and color correction
.jpg photos
Automatic masking
.ply Dense point cloud With holes
Multiscale model to model cloud evaluation
.obj High dense mesh
.E57 Vertex point cloud Without holes
.obj Mesh Decimation Alignment Scaling
.png Images for nail detection
GEOMETRIC AND TEXTURAL EXTRACTION
POINT CLOUD
250K FACES MESH
QUAD REMESH
QUAD WITH TEXTURE
NORMAL DISTANCE MEASUREMENT AGAINST A PLANE
MULTISCALE MODEL TO MODEL CLOUD
SURFACE CURVATURE ABSOLUTE CURVATURE QUADRATIC FITTING SEARCH RADIO 4
CURVATURE EVALUATION OF MESH SKELETON (WOOD WARPING)
BOW
CROOK
KINK
CUP
TWIST
TEXTURAL NAIL HOLE DETECTION
FACE : 0 [[0.6050, 0.9242], [0.4369, 0.4606], [0.4138, 0.05095]] Mask R-CNN Segmentation Output
Detections Stored as Element UV Coordinates
MASS EVALUATION
DIMENSIONS = 712x45x45mm VOLUME = 1441800 mm3 = 0.0014418 m3 SPECIES = European Redwood SPECIES DENSITY = 510 kg/m3 EXPECTED MASS = 0.735 kg ACTUAL MASS = 0.669 kg MASS RATIO = 0.91
DIMENSIONS = 954x45x45mm VOLUME = 1931850 mm3 = 0.00193185 m3 SPECIES = European Redwood SPECIES DENSITY = 510 kg/m3 EXPECTED MASS = 0.985 kg ACTUAL MASS = 0.538 kg MASS RATIO = 0.55
TEXTURE EVALUATION Texture Variation per element 4 faces per element
Filter Type
34%
Filter threshold based on luminance domain of 75% below average per item
Texture Variation Percentage
37.71
26.04
41.44
24.41
Saturation
18.01
19.41
20.13
9.39
Luminance
21.08
21.49
23.72
36.01
Low Texture Variation
High Texture Variation
8%
92%
TRACEABILITY
MS
UUID
Marking Options - Ink : Issues with Visibility / Application - Labels : Wasteful, Subject to Wear - Laser Etching : Need to Test
Dimensional and qualitative information
UI FOR ELEMENT INSPECTION
WEB INTERFACE
DESIGN DEMONSTRATOR
ELEMENT MATCHING PRE-SELECTION
BASED ON MATERIAL HEALTH
0-50%
DOWNCYCLED
50-70%
70-100%
CLADDING
STRUCTURAL
GENERATIVE DESIGN
Generative Design
Database
DESIGN - MATERIAL USAGE
Element dimensions have high flexibility once bridging connection points
Design information and specificity is embedded in the shape of the connectors between layers
ELEMENT MATCHING
Scenario A
Scenario B
DATASET ITEM MATCHES WITHIN TOLERANCE
CUTOFF ADDED BACK TO DATASET
Scenario C
[...]
[...]
ELEMENT CUT TO MATCH DESIGN
UNMATCHED ELEMENT WILL NEED NEW MATERIAL
MATCHING CRITERIA BASED ON NAIL POSITION Matched
X Connector
X
=
X Connector
x
X
Database Element
Target
OPTIMIZATION
RESULTS
# of solved connectors
85% of reused material From which only 10% requires labor
15% of new material Design data Database data Before
After
# of uncut and reusable material used
FABRICATION EFFICIENCY
STOCK USAGE BY RESULT
AVERAGE EQUIVALENT LINEAR WASTE USED : 65% WASTE : 8% REUSE : 27%
FABRICATION AUTOMATED CONNECTOR PRODUCTION
FABRICATION COMPUTER ASSISTED ASSEMBLY
CONNECTORS ID
TOLERANCE 10%
ORIGINAL LENGTH OF THE ITEM
STORAGE LOCATION
DESIGN LOCATION
P SIDE COLOR CODE
S SIDE COLOR CODE
FABRICATION COMPUTER ASSISTED ASSEMBLY
FOUND ITEM
LABOR NEE
EDED ITEM
NOT FOUND ITEM
FABRICATION COMPUTER ASSISTED ASSEMBLY
ASSEMBLED DEMONSTRATOR
Final assembly
Assembly Process
DEVIATION FROM ORIGINAL DESIGN Histogram
85% of pieces within 10 mm tolerance
Top View
Isometric View
ASSEMBLED DEMONSTRATOR
PILE OF TRASH
OUR DESIGN
PRIOR DEVELOPMENT
IDENTIFYING IMPERFECTIONS WITHIN THRESHOLD OF MESH CURVATURE
IMPERFECTIONS AS DESIGN OPPORTUNITY
AGGREGATION DESIGN WAVE FUNCTION COLLAPSE (WFC)
THIRD DESIGN DEMONSTRATOR
COMMERCIALIZATION
Matter Site is an innovative digital decision support for deconstruction and demolition. It employs an automatic digitization method for the near-end life cycle stage of a building and considers it as a source of high value assets.
Website: https://mattersite.co/ Blog: https://drivenbyvolumes.io/2021/03/18/driven-in cubation-platform-introduces-mattersite/#more-5 637
TEAM MEMBERS
External Admin Services [Management, Contracts, Legal Papers...]
Architect/Engineer [C&D waste technology specialist, European Market]
Sales [buyer domain expert]
Marketing [Communications and graphics]
Software development [Web App, Data Analysis, storage/security]
+2
Harvesting expert [Material recovery specialist, source domain expert]
Startup 140.600€
VALIDATION PRODUCT/MARKET FIT
BIG CIRCULAR ECONOMY NETWORK
INSIGHT CURRENT WORKFLOWS/PRODUCTION CHAIN
POTENTIAL PILOT COLLABORATION/ PARTNERSHIP
USER JOURNEY
SERVICE DURATION
ROADMAP
Incubation program started
Q4 - 2020
Initial industry collaboration/ Pilot
First Pilot Site scanning Hardware collab
Second Pilot Product fit evaluation. BM validation. Larger scale
Q1 - 2021
Q2 - 2021
Q3 - 2021
5 ongoing projects
3 ongoing projects
1st billable service implementation
Third Pilot Business model consolidation. Investors
Q3 - 2022
Q2 - 2022
Q1 - 2022
Q4 - 2021
TEAM MEMBERS 3 YEARS PROSPECTION [3 employees]
[6 employees]
[12 employees]
OPERATING COSTS
FINANCIAL PROJECTIONS 3 YEARS PROSPECTION
FINANCIAL PROJECTIONS 3 YEARS PROSPECTION
FINANCIAL PROJECTIONS 3 YEARS PROSPECTION
SEED CAPITAL INVESTORS
StartUp team
2nd round
1rst round
50% -12% (10-20%)
50%
18.2% +0.8%
30%
50%
20%
+20% 1yr
+20% 2yr
+20% 3yr
100% 4yr
CALL FOR PILOT MATERIAL
WHAT WE ARE LOOKING FOR
WEBSITE DESIGN