ground crack detection system - emsd
TRANSCRIPT
Proposal ID: EMS202002
E&M Inno-Portal Ref. No.: W-0238
Ground Crack Detection System
Prepared by:
Yolanda Au
Business Development Manager
Electrical and Mechanical Services Department
This document is strictly confidential and has been prepared by RaSpect Intelligence Inspection Limited and reviewed by Electrical and Mechanical Services
Department, solely for information use for the introducer of the said Inspection project and may not be taken away, reproduced, redistributed or passed
on, directly or indirectly, to any other person (whether within or outside your organization / firm) or published, in whole or in part, for any purpose.
The information contained in this document has not been independently verified. No representation or warranty expressed or implied is made as to, and
no reliance should be placed on, the fairness, accuracy, completeness or correctness of such information or opinions contained herein. The information
and opinion contained in this document are provided as of the date of this presentation, are subject to change without notice and will not be updated or
otherwise revised to reflect any developments , which may occur after the date of presentation. It is not the intention to provide, and you may not rely on
these materials as providing, a complete or comprehensive analysis of the company’s financial or trading position or prospects, and some of the
information is still in draft form. None of RaSpect Intelligence Inspection Limited or Electrical and Mechanical Services Department, nor any of their
directors, officers, employees, advisers or representatives shall have any liability whatsoever (in negligence or otherwise) for any loss howsoever arising
from any information contained or presented in this document or otherwise arising in connection with this document.
This document contains statements that reflect the company’s current beliefs and expectations about the futures as of the respective dates indicated
herein. These forward- looking statements are based on a number of assumptions , they are not a guarantee of future performance. Accordingly, you
should not place undue reliance on any forward- looking information. Each of the company and the Joint Bookrunners assumes no obligation to update
or otherwise revise these forward-looking statements for new information, events or circumstances that occur subsequent to such dates.
The distribution of this document in other jurisdictions may be restricted by law, and persons into whose possession this document comes should inform
themselves about, and observe, any such restrictions.
Disclaimer:
2Strictly Private & Confidential
Introduction
Strictly Private & Confidential3
Purpose - Support different level of ground crack detection of a minimum coverage of 15m x 15m
- Classify structural and non-structural crack
- Send out system alert when the monitoring parameters fall below threshold
Executive Summary
Strictly Private & Confidential
B. Monitoring Tools & Software Technology- Professional Grade Visual Camera
- Computer Vision Analytics System- LoRaWAN Data Streaming & Networking System- Realtime Feedback & Alert System
- Report & Data Storage Cloud Platform [https://www.raspect.co]
C. Team- Project Manager
- Software Architect
- Mechanical Engineers
- Artificial Intelligence and Deep learning Experts
A. Key DeliverablesPart 1 – Hardware
- 1A. Define Project Scope
- 1B. Preparation and Installation of Visual Camera
- 1C. LoRaWAN network installation
- 1D. Camera Adjustment and Establish Connectivity
Part 2 – AI Analytics Cloud Platform
- 2A. Visual Data Collection - 2B. Data Analysis through AI Analytics Platform
- 2C. AI Deep-learning Evolution
Part 3 – Continuous Monitoring and Alert System
- 3A. Data Streaming through LoRaWAN- 3B. Periodic Report Generation
- 3C. RaSpect Cloud-based Reporting Platform- 3D. Realtime Alert System
Others - Optional Features
- AI-powered Thermal Analytics
- True-scaled measurable 3D model reconstruction
4
Proposed Solution: Ground Crack Detection System
Strictly Private & Confidential5
Data Capture
To capture the visual appearance
of the targeted region by visual
camera (Camera model to-be-confirmed)
Data Analytics
To carry out defect identification &
classification, by AI Analytic Engine
Either on Cloud-based, or local
processor
Real time monitoring
To present and display the detection
results, by web portal and integrated dashboard
Data Streaming
To receive the captured image
from the network camera
Reporting & Alert
To notify and alert users when
critical defects are found, through
LoRaWAN communication protocol
AI Evolution Training
To conduct AI deep-learning to
improve the detection accuracy, by
accumulating annotated defects
Executive Summary
System Architecture Flow Chart
Scope of Work
Part 1 – Hardware1A. Define Project Scope
Discuss the site plan with relevant authorities before the commencement of operation, make sure the site environment is
favorable and safe to conduct the hardware installation; For the purpose of accurate result and alert notification, benchmark
for cracks detection and acceptance tolerance will be defined with the authorities.
1B. Preparation and Installation of Visual Camera
Visual Camera (To-be-confirmed) is installed above the targeted/potential ground crack area, ensure the camera is able to
capture the entire ground condition. Extra cameras will be installed when necessary.
1C. LoRaWAN network installation
LoRaWAN network will be installed on site and connection will be established according to Client requirement.
1D. Camera Angle/Position Adjustment and Establish Connectivity
Adjust camera angle or position if necessary, ensure the visual data for entire area of interest is clearly captured. Connect the
camera output to the analysis system and connect with LoRaWAN network .
Strictly Private & Confidential6
Scope of Work
Part 2 – AI Analytics Cloud Platform2A. Visual Data Collection
Connect Cameras to control terminal through wireless network, secure power supply for continuous operations. Cameras
capture a set of visual data once a day, or on the authorities’ requirements. The visual data are streamed to the AI
Analytics Engine for real-time data processing.
2B. Data Analysis through AI Analytics Platform
RaSpect AI Analytics Engine process the videos taken simultaneously on site. Identify crack defects through computer
vision technology. AI analytics engine will also analysis the severity of the cracks through AI engine training.
Strictly Private & Confidential7
(Ref. Case)
Analysis and Evaluation of Facade Condition
(Ref. Case)
Defects finding by using AI-Powered Analysis Algorithm
Scope of Work
Strictly Private & Confidential8
2C. AI Deep-learning Evolution
RaSpect AI Analytics Engine undergoes machine learning during the
operation. With the expansion of data sample, the reliability and
accuracy of the defect detection improve. Evolve into a rapid,
accurate and reliable detection system.
Benefits• Prevent catastrophic damage / accidents
• Diagnose ground cracks at early stage and predict time to failure
• Reduce unnecessary over-haul and maintenance costs
• Realtime monitoring and alert
Possible defects to be identified
• Cracks
• Honeycomb
• Delamination
• Rusting
• Staining
RaSpect Cloud-based Inspection Platform
Part 3 – Continuous Monitoring and Alert System
3A. Data streaming through LoRaWAN
All data assets posted and uploaded to custom-configured RaSpect cloud-based platform.
3B. Periodic Report Generation
A summary report is generated at pre-set intervals, summarizing the inspection findings. Identify cracks development trend,
perform predictive analytics. The materials can be reviewed to a custom-configured, RaSpect cloud-based platform.
3C. Data Visualization
All visual data sorted into folders by regions and timeframe. All recorded data are visualized in PDF catalogue
reference documents.
3D. Realtime Alert System
Prompt alert signal to person-in-charge when general aging and high-risk defects are detected. Real-time alert message
generated to Clients in email / sms or other forms.
Strictly Private & Confidential9
Scope of Work
10
Optional Features 1
11
Optional Features 1
Strictly Private & Confidential
(Ref. Case) Automated Thermal Anomalies Detected by
RaSpect AI-powered Façade Analysis Engine
Potential façade defects to be inspected
by visual & thermal:
- Cracks
- Stains
- Rusting
- Spalling
- Corrosion
- Scratch
- Water leakage
- Oxidation
- Coating Damage
Optional Features 1
Strictly Private & Confidential
13
(Ref. Case) Inspection Report summarized potential defect areas that automatically identified by RaSpect AI-powered Façade Analysis Engine
Optional Features 1
True-scaled Measurable 3D model
True-scaled measurable 3D digital model reconstruction of the pre-defined region, 3D model records the entire overview of the
targeted/potential crack area from time to time for defects trends comparison. On-request measurements of specified parameters. Level of
accuracy achieve ± 5cm.
Strictly Private & Confidential
(Ref. Case)
Photorealistic measurable 3D model: (Left) large scale 3D model; (Right) small object 3D model
Optional Features 2
Head Office
+852 59106300
+852 39607100
+86 13028850646
Phone
www.raspect.co
Website
16F, 700 Nathan Road, Kowloon,
HONG KONG.
HKAI Lab
6F, 10, Science Park W Ave,
Science Park, New Territories, HONG KONG.
Contact Us
Strictly Private & Confidential15