Proposal For EMSD iCremation Proposal Prepared by Wild

Proposal For EMSD i-Cremation Proposal Prepared by: Wild. Faces Technology Ltd 10 th Sept, 2019

EMSD – i-Cremation Background EMSD would like to use Artificial Intelligence (AI) technology to control the cremation process automatically Save energy to reduce cremation time Improve the efficiency of the process Reduce emission Reduce the work load of operator

EMSD – i-Cremation Wild. Faces’ AI Approaches: Understand pattern and apply AI to make whole cremation process more efficient • Making use of our AI technology to automate fairly manual process & to achieve energy saving by automating the processes Combination of multiple algorithms including patented technologies • • Alert operator which parts require higher air pressure Require images of different stages of the cremation • Including temperature, pressure, time, etc Step 1 – using Wild. Faces’ patented Non-motion Object Detection • • • Making use of existing HD CCTV camera at rear part of cremators Able to detect dark object in the moving flame Able to detect changes in the cremation process Step 2 – Colour Analysis • • Focus on the size/density of the dark areas Differentiate the location of dark areas (upper or lower) Step 3 – Interface our algorithms with existing Cremator Control Process

Cremation Optimisation • • Ensure whole object is completely cremated Black areas in image indicates parts that are not well combusted – • Temperature controlled to below 1000 o. C – • Require higher air pressure to increase combustion At Optimal temperature = 850 o. C Ideal burning time = 90 minutes – Max burning time = 180 minutes Black Areas 4

How it works – Step 1? Patented Non-motion Object Detection • • This is used as the framework of our software’s detection During combustion, detection of non-motion objects will be performed – • Raise alarm if it detects parts of the body is black – • Cremation residue will be detected whilst ignoring the flame This indicates it is not well cremated Further detection of black parts will be performed depending on the area size and location of the black parts Black Area detected at the top of the detection area Black Areas detected on the right side of the detection area 5

How it works – Step 2? Colour Analysis • • Can work with existing camera Detect the ‘black’ parts which are not completely combusted Depending on where the black areas are, which side of the air inlet will be used to enhance combustion Can indicate when to increase the combustion rate 6

How it works – Step 3? Synchronization of above with Cremator Control and Monitoring Software • • Patented Non-motion Object Detection is used to detect the changes during the cremation process Colour analysis is used to further detect the area size and location of the black residue Based on the above information, the temperature and combustion of the cremator is adjusted Detection will only stop until cremation process is completed Combination of all AI together Non-motion Object Detection Colour Analysis Communication with Cremator control and monitoring system Cremation Process Complete 7

Centralised Architecture + Cremation Detection Client Workstation Cremator Switch Events/Alarms Monitoring … 6 Unit Integration with cremator control and monitoring Event data Consolidation Video Analytics Central Server A consolidated view of 8 the events & a detailed report across all locations

Bill of Materials (BOM) and Roles & Responsibilities BOM By Wildfaces Analytics Software License (per camera basis) X Specialised Cameras X 1 Cremation Detection No. of camera Software License 6 Hardware • 1 x Server • 6 x Existing Cameras • 1 x 8 port POE Switch Customisation • Integration of software with cremator control and monitoring system • Training model for learning • • X X Ongoing Annual Maintenance -Software License Gold Support -Professional Services for Fine Tuning X X X LAN with cable all E&M work Wildfaces Professional Services: By End User X X X

Why choose us? Capabilities i. Omniscient/Wild. Faces # of Commercial Experience in Analytics Track Records 18 years ? Successful smart City projects deployed in close to 60 countries ? Patents in Crowded and Complex scenes Comprehensive Capabilities 50+ international patents ? Multi-sensory Analytics (video, Sound and Smell) Plenty of Smart city use cases ? Many – see our partial reference list ? Yes – to many… ? Solved Customer Problems – Use Case Basis Approaches References Design Assistance offered to Customers Other Suppliers ? www. i. Omniscient. com | www. wildfaces. ai 10

World Leader In A. I. Deployed in close to 60 countries Comprehensive solutions for 30+ industries Multi-sensor analytics Global Support 50+ Patents Excels in extreme crowds Flexible architecture Expert in complex scenarios Big Data Analytics 11

The World Leader In Artificial Intelligence Bringing Its Advanced Technologies At Affordable Prices and Risk Free Thank You

References HONG KONG ‘THE ONE’ MALL HONG KONG POLICE HONG KONG MTR HONG KONG Pi. K UK PRISON SUNTEC SINGAPORE SMART CITY IRAQI SMART CITY DUBAI MALL DUBAI INVESTMENT PARK DUBAI ELECTRICITY & WATER AUTHORITY PULLMAN HOTEL CUSTOMS, KIND FAUD CAUSEWAY BUDAIYA POLICE …and more Please find below for the complete worldwide project reference list: https: //www. dropbox. com/s/sdore 0 bv 3 beutpl/ Wild. Facesi-Omniscient%20 References. pdf? dl=0