Case Studies

Real World Applications

STech.ai has made its mark in the industry and is proud to claim to be engaged with the public and private sector. With due accreditation to the agreements that have been made with these institutions and otherwise consideration of privacy and security of data, STech.ai cannot reveal the names of its clients. However, their experiences are narrated in the form of the challenges they faced and our aptly responsive solutions.
Stated below is a brief summary of what we have been up to lately.

1

Smart Transportation

Traffic Analytics for a Middle Eastern city with an approximate population of 3.5 million

Challenge: The city with an increasing population faces an even rapidly increasing traffic congestion. Since it is affected by tourism and religious services, observation suggests crucial concerns on traffic patterns. This in turn affects planning and management which is rather unreliable and lacks far reaching technological advancement.

Response: STech.ai deployed its vehicle counting and classification system which was integrated with the existing infrastructure. With the implementation, the planners and managers could now easily count and classify the vehicles to make valuable planning and implementation decisions. STech.ai's system provided them with the real time data which gathered for them, their short term and historical data, valuable for growth analysis. Among several other benefits, reduction of human cost of traffic analysis, collection and analysis of data, management and automation of short-term tasks (e.g. changing traffic lights)top the chart.

2
License Plate Recognition System (ALPR) for a capital city in Asia with a population of almost a million

Challenge: License Plate Recognition has become a necessity with the modern day rising security and safety challenges. In a city with more than sixty types of standard and non-standard license plates, it became almost humanly impossible to recognize and identify the licenses using CCTV footage. Even in most favorable circumstances, it is likely to have an outcome full of errors.

Response: STech.ai masters customization of solutions, hence for a project like the one mentioned, STech.ai has provided customized Automated License Plate Recognition (ALPR) to cater to their requirements. With a wide training network and several thousand localized models, STech.ai’s ALPR results in almost 80% accuracy, leading the charts in the industry. The potential for accuracy is limitless as increased data processing leads to increased accuracy. As a consequence, traffic counting and classification has also been added to the system upon client’s request for the machine to process and analyze more data.

3
Validation of the collected toll for an Asian country’s National Highway Authority

Challenge: National Highway Authority owns all the tolling related processes and the stations. In such scenarios, a detailed report of the performance of the private contractors at various stations is required for qualitative and quantitative analysis. Data is also required for analysis of traffic flow and patterns to be identified for growth mechanisms. These processes require ample human resources which causes a financial crunch and are too time consuming causing waste of the ability of the human resource. Margin of error is also great and there is always room for bias.

Response: Lately, CCTVs have been installed in several tolling stations. STech.ai, using its ability to integrate with existing infrastructure has made full use of the CCTVs and has provided a highly accurate and real-time data collecting solution for revenue validation, vehicle type classification and count. This service is a pioneer in itself, providing in black and white the data for performance evaluation, easier management, planning and scalability. Expenses cut down, error minimization, real-time and reliable technology, incorporation with existing infrastructure are key features of the deployed solution.

 

4
Parking Management System with ALPR for a humongous shopping mall in an Asian city of 1 million

Challenge: The mall is a multi use facility which is a 36-floor hotel, numerous floors of office and residential facility, along with a 5-floor shopping plaza. The problem they faced was to cater to the vehicle parking for consumers of all these facilities. It required dynamic decision making, deployment of human resources whilst providing for their training and expertise. On top of it, providing safety, security and accurate divergence of traffic in the appropriate direction is a cumbersome task.

Response: STech.ai came to the rescue for the facility to aptly manage their parking system, complying to the needs and wants  of the concerned party. STech.ai provided them with:

  • An efficient identification system for registered and unregistered vehicles within the premise
  • Control and segregation of traffic according to their type (residential, office, mall or hotel)
  • An identification mechanism for the vehicles that take up parking spaces from nearby areas
  • A real time identification mechanism for the available parking spots in the various parking areas of the facility
5
Traffic Analytics for a Middle Eastern city with a population of almost 1 and a half million

Challenge: This Middle Eastern city has constantly been on the rise ever since 1970. It has become one of the greatest business arenas and the flow of traffic continues over time as vehicles from neighbouring cities continue to move to and fro. Such variation in traffic has surfaced as a major traffic management challenge for the Road Transport Authority. However, the city is always open to incorporate the innovation and improvement of technology in automation and machine learning. Hence, the Authority requires immediate responsiveness to their prevailing situation.
 
Response: The Road Transport Authority is committed to innovative solutions, which is why STech.ai was able to provide for them, according to their desires and better than the currently installed mechanism, its Vehicle Detection, Classification and Analytics system. This system made the collection of historical data and analysis 85% faster than the previously installed system, that too all in REAL-TIME. STech.ai has provided the Road Transport Authority with a modern solution to their problem which is not only reliable but also benefits them with accurate data for further deep analysis  and user friendly for its operators. 

 

6
Vehicle Identification for a large, gated Residential Facility in a major Asian City

Challenge: Vehicle theft is a common vice in the neighbourhood for this Asian city. Gating, fencing, security checkpoints, installing CCTVs or even using RFIDs is not a great obstacle in the theft of the automobile. The possibility of the vehicle leaving the premises despite a RFID tag is considerable as the owner’s recognition is not a compulsion of the movement of the vehicle. Hence arose the need for a secure mechanism for the vehicles to move around and be parked freely.
 
Response: STech.ai’s easy integratability came to the rescue for the facility by providing the Automated License Plate Recognition (ALPR) System using their existing security surveillance cameras. The ALPR system provided access control with an identity verification in real-time using the resident database. The system allows an SMS to be generated to the owner whenever the vehicle is detected at the entry or exit point. As a way forward, it could also be integrated with the facial recognition system for the automated opening and closing of the gates upon the identification and verification of the owner/resident in the vehicle with the license plate tag.

 

7

Facial Recognition

Access Control System for Manufacturing Facilities using Facial Recognition

Challenge: A major sports wear brand has its manufacturing facility employing around 24,000 workers at its four locations. At one such location, on average, around 6,500 employees are working at shift switches within twenty minutes, which means that the shift causes slow moving, long queues. The management of entry and exit is done using RFID and biometric signatures which is linked with their ERP for their record. This situation created a time management crisis for the manufacturing facility.

Response: The challenge presented to us by the company was to provide a mechanism for quick recognition and verification of the employee, record and report system linked with ERP and lock/unlock the turnstile gate. They wanted this done within a fraction of a second. The efficient and responsive solution to their existing problem was STech.ai’s Facial Detection and Recognition system, which allowed an employee to be detected, recognized, verified and enter or exit through the turnstile, all within a second and through either of the four premises. It is a complete package, which is not only efficient and responsive but is accurate, scalable and cost-effective.

8

People

Crowd Analysis for a Middle Eastern city catering religious congregations

Challenge: The city is a sacred site for the global muslim population. There are massive gatherings on religious occasions which turn in a lot of revenue for the cities and the country, altogether. Such gatherings often cause collisions and stampede. Security is also a major concern in such areas. The challenge presented was to analyse the flow of crowd and predict any circumstances that could possibly be avoided to refrain from an unfortunate outcome.
 
Response: STech.ai was invited to assist the city with its overpowering problems. As a result, STech.ai provided them with the crowd analytic platform which assisted them to predict and prevent any unfortunate event that could ruin the essence of the gathering as well as ensure the swift and smooth flow of human traffic to avoid any inconvenience. 

 

9

Agriculture

Climate Smart Farming for farm owners in an agriculture prone province of a South Asian country

Challenge: A common observation in the area was to lose crops due to reasons accredited to the climate and weather changes. The surveys suggested that the farmer’s crop would get destroyed at an early stage because of the unprecedented changes in the weather and climate, attack by pests and insects or poor management.
 
Response: STech.ai responded to the crisis with its MechaSapien to deploy sensors in the field and allowed the farmer to respond to the predicted changes in the weather and climate, identify the part of the crop attacked by pests and insects. The farmer was facilitated by providing him with a user friendly interface which is easy to interpret and understand.

 

10
Agricultural Land Assessment for private landowners in a rural city of a South Asian country

Challenge: The city is responsible for most of what crops in the country and several landowners suffer with similar problems. The land guarantees a progressive yield but the crop growth is uneven at various parts of the land. The use of fertilizer is a must and henceforth raises the question of what criteria is followed for sprinkling the fertilizer on the land. The problem faced by most of the farmers is to spend too much on the fertilizer to ensure a stable crop.
 
Response: STech.ai’s capability was able to facilitate the farmers by providing them with a fertilizer application map which gave the farmers a definite idea of which part of the land required fertilizer and which could do without it. A survey of the land was done by an autonomous flight of a drone which captured the video of the land and analyzed the quality. Further, the video was processed to generate heat maps which clearly identified the areas which required extensive fertilizer and which were fertile enough. This assistance provided the farmer enough information to be able to reduce the fertilizer cost by 20%. 

 

11

Video

Revenue Verification for a private contractor to the National Highway Authority of an Asian country

Challenge: A private contractor is in charge of managing the toll collection and tolling stations. The contract for this task is given for two years by the National Highway Authority through a scrutinized selection process. The contractors are often seen to suffer losses due to several problems which include  incorrect estimation of traffic and revenue, which is an entirely human dependant process.

Response: We are proud to state that automated technology at tolling stations in the discussed country has been introduced by STech.ai. STech.ai has provided them with a digital counter unity that is deployed at various stations. The portable unit has the capacity of storing the data of a minimum of seven days to identify any changes in the vehicle pattern and then collects and analyzes this data which is beneficial for the contractor in their future endeavours. The data provided by the unit is multiple times accurate than the manually gathered data, hence the planning and decision making is more effective.