AI Case Studies

Vehicle Detection, Analytics and Classification

Traffic Counting & Classification – Major Middle Eastern city of 3.5 million people

Problem: Traffic congestion is high on one of the main thoroughfares in the city. Tourism and very large religious congregations cause critical spikes in traffic patterns. Management and planning relies on relatively unreliable and low-tech data extrapolation.

Solution: CRE8 implemented a vehicle counting and classification system to existing infrastructure. City planners and managers now have accurate vehicle count and classification (type of vehicles, etc.) information to make real-time, short-term and historical data for long-term growth decisions. CRE8 Systems helped provide accurate real-time and historical information with analysis. Eliminated the human cost of traffic survey and analysis, reduced the human cost of data analysis, reduced planning and management time and assisted in automating short-term tasks, such as traffic light changes, etc.

Automated License Plate Recognition (ALPR) – Capital city of 1 million in Asia

Problem: As part of a citywide safety and security initiative, License Plate Recognition became a key requirement. With over sixty different types of non-standard and customized license plates in circulation, it turns into an unaddressable problem using only human resources and CCTV footage. At the very least, It becomes highly error-prone.

Solution: CRE8 adapts and customizes it’s License Plate/Tag Recognition (LPR) technology to account for the specific needs of the non-standard environment. CRE8 trains the learning system with over ten thousand localized models and scenarios. Resulting in an 80% accuracy rate which is extremely high in the industry. This number keeps growing as more and more data is analyzed and learned by the system. Based on these results, the City has also added Traffic Counting and Classification to the mix, giving our systems even more data and input.

Revenue Verification – Private contractor to a country’s National Highway Authority in Asia

Problem: A major private contractor to the National Highway Authority of a country in Asia is tasked with the management of toll collection stations. Toll collection stations are offered to contractors for 2 years based on various bidding, estimation and selection processes. It is common for these contractors to end up in losses due to incorrect vehicular and toll collection estimates. The estimation is a completely manual process and based on a very small maple size.

Solution: There are next to no IT services available at many toll collection stations throughout the country. CRE8 provides a portable traffic counter unit that is deployed at various stations. The units record data for a minimum of seven days to account for any vehicle pattern aberrations and then collects and reports the analyzed data to the private contractor for their bid renewals. Since this data is magnitudes more accurate than previous estimations, it gives the private contractor a very tangible and large competitive advantage.

Toll Collection Validation - Asian country’s National Highway Authority

Problem: The National Highway Authority of a country in Asia owns all the toll collection stations. Based on historical data, the Highway Authority requires an independent report of station operator performance (private contractors). In addition, the Highway Authority requires data on vehicular flow and congestion patterns to better plan for expansion and enhanced services for toll collection stations. All this information requires costly human resources and estimation tools are highly inaccurate and there is additional room for bias in the data.

Solution: There is a healthy and growing percentage of toll collection stations that have CCTV infrastructure. CRE8 systems are incorporated to provide an accurate and real-time assessment of toll revenue collection, vehicle count and type. For the first time the Highway Authority has an accurate gauge of revenue flow and bankable information and reporting to aid in private contractor selection, management, planning and scalability of toll collection stations. Expensive and error prone human resource cost is eliminated and replaced with real-time, accurate and reliable technology, using existing CCTV infrastructure.

ALPR Based Parking Management System – 5-floor shopping mall in an Asian city of 1 million

Problem: Managing parking for a mixed use development that is a 36-floor hotel, 23-floor office and residential towers and a 5-floor shopping mall is no small feat. Parking management in such a highly dynamic environment, especially with the shopping mall traffic variability, traditionally requires a lot of human resources, training and expertise. Add to this layers of public safety, security and segregating vehicular traffic to the right areas (residential, office, hotel, shopper) is what keeps management staff awake at night.

Solution: Working with the requirements and wish lists of the property management, CRE8 solutions fundamentally address these key functions for three different parking areas:

• Identification of registered & unregistered vehicles with property management
• Segregation and control of vehicular traffic based on type: residential, office, hotel or mall
• Identification of idle vehicles taking parking spaces from nearby retail, offices, etc.
• Real-time identification of available parking spaces in the different parking areas

CRE8’s packaged and automated solution addresses all Client needs. There is also real-time management alerting system on aberrations and discrepancies. Management has complete access to historical data for better planning and service enhancements.

Traffic Volume Analysis – Major Middle Eastern city of 1.4 million

Problem: A major Middle Eastern city exploded on the world scene in the 1970s and has not stopped growing since. One of the major business hubs in the world, this city grows and shrinks by the hour due to commuting traffic from and adjacent city, another world business hub. Due to the vast variations in a day to vehicular traffic, the major roadways are a nightmare for the Road Transport Authority. This city embraces technology as a way to continuously move forward and knows the solution lies in automation and machine intelligence. The Transport Authority needs answer and it needs them fast.

Solution: Working with the Road Transport Authority and accounting for the wishlist, after a competitor’s installation, CRE8 implements its Vehicle Detection, Analytics and Classification system. In addition to real-time data of volume study, historical data collection and analysis is provided 85% faster than the competitor. With CRE8, the Road Transport Authority opted for a replacement technology that better serves their requirements while providing accurate and reliable data analysis for traffic management and planning.

Facial Recognition

Manufacturing Facilities – Facial Recognition Based Access Control

Problem: A large manufacturing facility for a global sporting brand has 24,000 factory employees working at four different locations. On average a single location has around 6,500 employees trying to switch shifts in a span of twenty minutes. The result is long and slow moving queues. Managing the entry and exit using RFID and biometric signatures that are integrated with their Enterprise Resource Planning (ERP) for payroll and other administrative records. All this to happen for 6,500 employees within twenty minutes was simply not happening.

Solution: The manufacturing facilities management challenged CRE8 to come up with a solution that would recognize an employee, verify, record and report via the ERP, unlock/lock a turnstile gate, within one second. Their requirement was a swift mechanism which can process the entry or exit of an employee in less than a second. They also required the integration of such a system with turnstile gates. CRE8 accomplished just that with our Facial Detection and Recognition: an employee/second entering or exiting any of the four premises, while satisfying all the core and ancillary requirements. Accurate, efficient, scalable and cost effective.

Residents Vehicle Identification – Very large, gated community in Asian city

Problem: Automobile theft becomes an issue for a large neighborhood in a major Asian city. Despite fencing/gating the entire boundary of the property, posting check points at entrance/exit points, instituting RFID based tags and adding CCTV, the issue has been marginally mitigated. Even with RFID, there is no console that matches the vehicle to the owner/resident and with this amoly, the vehicles are allowed to pass.

Solution: Using existing CCTV infrastructure, CRE8 implemented the Automated License Plate Recognition (ALPR) system is put in place. In real-time, the ALPR accesses and confirms the identity of the owner through the resident database. As a redundant security measure, an SMS is generated to the owner of the vehicle as ALPR detects it at entry and exit points. A future enhancement of the system is in the works where the Facial Recognition system works with ALPR to automatically open the gates on positive facial recognition match of the owner or residents and the license tag and vehicle description.