AI determines the severity of damage and provides rating automatically. Rapid image recognition is able to label potentially dangerous building defects, document their location and range and analyses their severity.
<img src="/uploads/visual_analysis_e33fcd7b36.jpg" alt="" width="100%" />

Visual Analysis

The data analysis is based on the following Codes and Standards -
<ol>
<li style="text-align: justify;">
Issued by Building Departments
<ul>
<li><a title="Code of Practice for The Mandatory BUILDING Inspection Scheme and The Mandatory WINDOW Inspection Scheme 2012" href="https://www.bd.gov.hk/doc/en/resources/codes-and-references/code-and-design-manuals/CoP_MBIS_MWISe.pdf" target="_blank" rel="noopener">Code of Practice for The Mandatory Building Inspection Scheme and the Mandatory Window Inspection Scheme, 2012</a></li>
<li><a href="https://www.housingauthority.gov.hk/common/pdf/business-partnerships/resources/concrete.pdf">Code of Practice for Structural Use of Concrete, 2013</a></li>
<li><a href="https://www.bd.gov.hk/en/resources/codes-and-references/codes-and-design-manuals/bmg.html">Building Maintenance Guidebook, 2002</a></li>
<li><a href="https://www.bd.gov.hk/doc/en/resources/pamphlets-and-videos/IYBS_e.pdf">Is your Building Safe? A quick guide to how to check it, 2014</a></li>
</ul>
</li>
<li style="text-align: justify;">
Issued by The Hong Kong Institute of Surveyors (HKIS)
<ol style="list-style-type: lower-alpha;">
<li style="text-align: justify;"><a href="https://www.hkis.org.hk/archive/materials/category/20160927174934.0.pdf">Professional Guide to Building Inspections, 2013</a></li>
<li style="text-align: justify;"><a href="https://www.hkis.org.hk/archive/materials/category/20160816141943.0.pdf">Professional Guide to Water Seepage, Investigation, Diagnosis, Testing &amp; Reporting in Residential Buildings, Building Surveying Division, 2014</a></li>
</ol>
</li>
</ol>

Data analysis is based on codes and standards issued by the Government & HKIS

<ol>
<li>
For reinforced concrete
 A thermal image can show the presence of moisture or delamination below a surface because of properties such as heat capacitance. Detection of thermal anomalies related to common concrete problems: e.g. voids, delamination, moisture and water leakage. 
<img src="/uploads/thermal_analysis_1_58b561999a.jpg" alt="" width="100%" />
</li>
<li>
For curtain wall Air leakage in aluminum panel joints caused cold air leak in from outdoor. Color difference shows in the thermal image.
Detection of thermal anomalies related to air leakage in aluminium panels joints. 
<img src="/uploads/thermal_analysis_2_9202e1d7b3.jpg" alt="" width="100%" />
</li>
</ol>

Thermal Analysis

Our AI Cloud Engine (ACE) is able to identify defects on different fac&ccedil;de construction materials 
<ol>
<li class="p1">Reinforced Concrete Crack, Stain, Delamination, Rust, Peeling, Sagging, Blistering, Fading, Moisture and water leakage etc. </li>
<li class="p1">
Curtain Wall Stain, Scratch, Missing panel, Abnormal panel, Chipping, Missing cladding, Missing sealant, Sealant deterioration, Bubbling sealant, Air leakage in aluminium panel joints etc.
</li>
</ol>

Recognitions for over 50 types of defects

The severity of all detected defects and thermal anomalies are assessed through a thorough micro inspection which scores the defect status as &lsquo;Cosmetic&rsquo;, &lsquo;Minor&rsquo;, 'Moderate', 'Major' or &lsquo;Critical&rsquo;.

Classification of the defect in 5 impact severity levels

Our AI Cloud Engine (ACE) is able to identify defects on reinforced concrete, curtain wall, cladding, bricks &amp; tiles.

Comprehensive inspection solution on different construction materials

large_key-visual-ai-powered-defect-detection-and-analysis.png

medium_key-visual-ai-powered-defect-detection-and-analysis.png

small_key-visual-ai-powered-defect-detection-and-analysis.png

thumbnail_key-visual-ai-powered-defect-detection-and-analysis.png

key-visual-ai-powered-defect-detection-and-analysis.png

SUBSCRIBE

GET A DEMO

ACE - Façade | Product

AI-powered
Defect Detection
& Analytics

AI-powered Defect Detection & Analysis

Fix or remove incorrect, corrupted, incorrectly formatted, duplicate, or incomplete data within a dataset.&nbsp;
<img src="/uploads/data_cleansing_1_40c26c0b4f.jpg" alt="" width="100%" />
Minimize the chance of containing duplicated or mislabeled data while combining multiple data sources. Unique identification of the defect. The defect is uniquely identified across multiple data snapshots.
<img src="/uploads/data_cleansing_2_c852d93792.jpg" alt="" width="100%" />

Data Cleansing

Estimate the parameters of the lens and image sensor, and adopt the coefficients for lens distortion correction.

Geometric Camera Calibration

This is an offset calibration usually performed at power up, when the camera changes temperature, and periodically during operation. This calibration compensates for certain errors that build up during camera operation. It requires a shutter or similar uniform temperature device to cover the camera field of view.

Flat Field Correction (FFC)

This is a stored FFC that can be used to provide a good image when using a shutter is not practical. It can be used with advanced processing for completely shutterless operation, or with a shutter for optimum operation.

Non-Volatile Flat Field Correction (NVFFC)

<ol>
<li>The data is categorized according to its level of sensitivity or value using labels. These labels are attached as visual markings and embedded into the metadata of the file.</li>
<li>Different building element features are extracted for segmentation. 
<img src="/uploads/data_classification_a5810e0baa.jpg" alt="" width="100%" />
</li>
</ol>

Data Classification

This calibration can compensate for image effects caused by the changing temperature of large lenses or other optical components. It may also help with effects from heat sources in camera housings.

Supplemental FFC

<ol>
<li>Finding the exact geometric location of these unmanned aerial vehicles is important.</li>
<li>Two estimation methods are proposed to solve the localization problem. The first scheme localizes the projection of the target object on the horizontal plane and estimates the altitude later. 
<img src="/uploads/data_localization_1_00772bbc2f.jpg" alt="" width="100%" />
The second scheme estimates the location of the target object directly in three-dimensional space. Data are categorized in a grid system for easier retrieval. 
<img src="/uploads/data_localization_2_0213172c06.jpg" alt="" width="100%" />
</li>
</ol>

Data Localization

This calibration may be used to optimize or refresh measurement performance on those cameras with Spotmeter or Image Metric optional features.

Radiometric Calibration

medium_key-visual-automated-data-processing-and-calibration.png

small_key-visual-automated-data-processing-and-calibration.png

thumbnail_key-visual-automated-data-processing-and-calibration.png

key-visual-automated-data-processing-and-calibration.png

Automated 
Data Processing 
& Calibration

Automated Data Processing & Calibration

Our AI Cloud Engine (ACE) is able to
<ol>
<li>Analyse large &amp; varied sets of data, uncover information including but not limited to hidden patterns and correlations.</li>
<li>Analyze current &amp; historical facts, detect faults at their early stage, identify faults and predict the time to failure.&nbsp;</li>
</ol>

Visualize the development of defect breakdown

<ol>
<li>Our predictive analysis generates a maintenance schedule based on the predicted development of defect breakdown.</li>
<li>Allows better planning in building maintenance or renovation works, preventing breakdown and accidents and reducing maintenance cost.</li>
</ol>

Maintenance Scheduling

medium_key-visual-big-data-predictive-analysis.png

small_key-visual-big-data-predictive-analysis.png

thumbnail_key-visual-big-data-predictive-analysis.png

key-visual-big-data-predictive-analysis.png

Big Data 
Predictive 
Analytics

Big Data Predictive Analytics

Our AI Cloud Engine (ACE) is able to interpret data with details from the visual image overlaid on the thermal data stream.
<img src="/uploads/data_categorization_11cd549c0d.jpg" alt="" width="100%" />

Categorize visual & thermal data captured by the dual camera in one snap.

medium_key-visual-visual-and-thermal-data-comparison.png

small_key-visual-visual-and-thermal-data-comparison.png

thumbnail_key-visual-visual-and-thermal-data-comparison.png

key-visual-visual-and-thermal-data-comparison.png

Visual & Thermal 
Data Comparison

Visual & Thermal Data Comparison

<ol>
<li>Executive summary and defect summary are provided to show a general overview of the inspection result.&nbsp;</li>
<li>Conclusion and recommendation is provided to conclude inspection findings and suggesting follow-up action.</li>
</ol>

Overview of the latest status of the Façade

The chart displays the following statistics with the selection of one specific elevation or the whole building:
<ol style="list-style-type: lower-alpha;">
<li>Total number of defects</li>
<li>Number of defects with the different severity levels</li>
<li>Number of defects of different defect type&nbsp;</li>
</ol>

Chart displays statistics of defects

<ol>
<li class="p1">Integrated with grid system for dimensions &amp; floors annotation</li>
<li class="p1">Display defects with different severity levels &amp; types</li>
<li class="p1">Select defect and data from 2D drawings for a detailed display</li>
</ol>

Visualization of defect location on 2D drawings 

medium_key-visual-building-facade-performance-dashboard.png

small_key-visual-building-facade-performance-dashboard.png

thumbnail_key-visual-building-facade-performance-dashboard.png

key-visual-building-facade-performance-dashboard.png

Building Façade 
Performance 
Dashboard

Building Façade Performance Dashboard

Our interface allow user to view rendered 3D Photogrammetry model with zoom, pan and rotation tools.

Model viewing interface for 3D Photogrammetry Model exploration

Our interface allow user to make measurement by selecting 2 points from the model.&nbsp;

Make measurements on model

<ol>
<li>Select models from timeline for comparison</li>
<li>Model overlaid function</li>
</ol>

Time-comparison of models with different data collection time

Our 3D Photogrammetry Model is in xxx format. User can download and view model on other 3D model viewers.

Download 3D Photogrammetry Model
(add-on service)

medium_key-visual-3d-photogrammetry-model.png

small_key-visual-3d-photogrammetry-model.png

thumbnail_key-visual-3d-photogrammetry-model.png

key-visual-3d-photogrammetry-model.png

3D 
Photogrammetry 
Model

3D Photogrammetry Model

&nbsp;
Follow us for our latest updates and subscribe to our LinkedIn page.

Ensure your assets' 
Safety, Serviceability and Sustainability with RaSpect.

CONTACT US

large_background-contact-us.jpg

medium_background-contact-us.jpg

small_background-contact-us.jpg

thumbnail_background-contact-us.jpg

background-contact-us.jpg

6900070091543719936

6888776279726931968

6873902845083693056

6862334655379709952

6851455917171601408

6830853396417912832

6825470055510347777

6820235588940165120

6799942072712482816

News

Building Predictive AI for Sustainability.
Stay current with our latest insights.