Securing a national border requires a delicate balance of effectiveness and efficiency. A border must be protected to block illegal contraband, smugglers and more. Travelers must be verified, and passports and other travel documents endorsed. Cargo needs to be inspected and compared against its paperwork. These duties must be undertaken with extreme care and attention. Yet it must also be done speedily, so that traffic — and commerce — continue to flow.

What’s more, there are further layers of complication. At land checkpoints, vehicles arrive in various shapes and sizes — cars, motorbikes, buses and trucks — and each needs to be handled differently. Also, each category (typically referred to as a “conveyance”) can include additional subcategories of vehicle types. For example, truck conveyances include oversized trucks, trucks carrying livestock and trucks carrying building materials.

For checkpoint agencies, achieving this delicate balance in real time has been difficult, and for several reasons. For one, traffic patterns at a border can be affected by factors including the weather and time of day. For another, detecting actual violations can be tricky. Criminals constantly work to stay one step ahead of border officers, and their illegal subterfuges can be ingenious and difficult to uncover.

91% of the world’s population are living in countries with border restrictions due to the COVID-19 crisis.

Source: Pew Research Center April 1, 2020

Another challenge is availability of resources. When unexpected traffic volumes may require additional staff, getting these personnel at the last minute is typically far from easy. Yet another challenge is aging infrastructure. On-land checkpoints, bridges, roadways and kiosks operate 24x7, and as a result are prone to heavy wear and tear. Repairing or replacing them is often postponed to avoid interrupting day-to-day operations.

A wide range of technology solutions are meeting these challenges. This paper explores those technologies, including a smart border checkpoint system developed by DXC Technology for one of the busiest checkpoint agencies in Asia. To stay secure and keep the traffic moving, today’s smart checkpoints must integrate artificial intelligence (AI), video and predictive analytics, internet of things (IoT) sensors, smart kiosks, drones, mobile apps and next-generation infrastructure. These innovations will enable checkpoints to achieve that delicate balance with only minimal human intervention.

The many challenges facing border agencies

Every border agency, no matter where it’s located in the world, strives to achieve a harmony between two vital but often conflicting goals: one, to ensure the highest possible levels of border security; and two, to move travelers along as quickly and efficiently as possible. 

This balancing act can present border agencies with serious challenges. On one hand, ensuring security takes time; on the other, delays can be extremely costly. Security is time-consuming because documents, vehicles and cargo all need to be carefully checked; travelers’ identities need to be accurately verified; and rules and regulations need to be scrupulously followed. Yet if the process takes too much time, a border agency can not only frustrate travelers, but also run up huge costs. 

$3.4B is the estimated annual cost of border-crossing delays at the checkpoint between San Diego, Calif., and Tijuana, Mexico.

Source: Feb. 27, 2021

For example, at the border crossing between San Diego, Calif., and Tijuana, Mexico, a single year’s worth of delays recently cost the region some $3.4 billion in lost economic output and 88,000 jobs. With traffic jams at this one border crossing running up to 10 hours, the situation also hurts the environment. Those idling engines waste literally thousands of gallons of fuel and spew tons of pollutants into the air. 

These problems are further exacerbated by the sheer number of people and vehicles crossing a border. At the world’s busiest land checkpoint, the ports linking mainland China and northern Macau, border agents process 134 million travelers a year. That works out to an average of about 367,000 people — more than the entire population of Iceland — being processed every day.

Another growing problem at many checkpoints is the space available for the volume of vehicular traffic. Cars and motorbikes may need to navigate across high-volume but narrow roadways. Traffic jams at some border crossings quickly pass from minutes to hours. Border checkpoints also demand significant land, with protected perimeters. Buses need plenty of room to park, so that their passengers can disembark and be cleared through the immigration halls. And cargo trucks require massive parking bays where they can be inspected. In addition, drivers of both buses and trucks require enough room within a checkpoint facility to execute turns, reversals and other maneuvers.

Further compounding these challenges, border agencies must also contend with numerous attempts at smuggling and illegal border crossings. X-ray and radiation scanners can help, as can staff and canines that have been trained to detect suspicious behavior and read the body language of smugglers. But the numbers and tactics remain daunting. For example, in 2019 checkpoint officers in Singapore discovered that a truck purportedly carrying a load of concrete blocks was actually trying to smuggle 12,500 cartons of untaxed cigarettes, all of them carefully concealed inside the concrete blocks. That same year, officers at the Singapore checkpoint stopped a man from illegally importing four kittens. The would-be smuggler was uncovered when the tiny cats were heard mewing from inside his pants.

10.6M consignments, containers and parcels cleared in a single year at the overland checkpoints between Malaysia and Singapore — representing 300,000 more cargos than were cleared in the previous year.

Source: Singapore Immigration and Checkpoints Authority

COVID-19 has introduced unique challenges, as well, and many border agencies have responded with new checkpoint measures. For example, in March 2020 Malaysia issued a 2-week movement control order that, among other restrictions, stopped the entry of all tourists and foreign visitors into the country. Similarly, in April 2020 Malaysia Immigration reduced one of its two checkpoint operations from 24 hours a day to just 12.

In addition, border agencies endeavor to meet their own targets, known as key performance indicators (KPIs), for customer satisfaction, often measured by onsite surveys. Inefficiencies at border crossings result in delays, which in turn create dissatisfaction among locals and tourists alike. This is likely to be reflected on the surveys, causing missed KPIs and declining satisfaction scores.

Creating a checkpoint prototype

To begin developing a next-generation checkpoint system, DXC Technology joined forces with one of the busiest checkpoint agencies in Asia. Working together, DXC’s Digital Innovation Lab and the agency co-created a prototype digital simulation of the checkpoint in just 5 weeks. Essentially, it’s a system that lets agents conduct virtual what-if scenarios by adjusting variables such as types of vehicles, duration of inspections, closures of specific roads and weather conditions. The system can model various scenarios on the ground, allowing for throughput analysis. Users can also view the simulation in both 2D and 3D. For example, the simulation can show whether opening an additional lane would relieve a traffic jam of trucks.

To develop the prototype, DXC first conducted on-the-ground research, which enabled the team to understand the daily problems faced in the checkpoints for the simulation, and to make an accurate 3D model, essentially a twin in digital form. The resulting system can now be used for a variety of planning activities.

To do this, the system employs several approaches of simulation modeling. These include: 

  • Spatial analysis, which is used to determine the effects of changing physical factors such as ramps, lanes and checking stations.
  • Flow-based analysis, used to account for delays caused by heavy rain, fog or even smoke from agricultural fires.
  • Monte Carlo simulations; here, a model for estimating the possible outcomes of an uncertain event is used to apply statistics that compare the time differences among various scenarios.  

The simulation prototype highlights new features likely to appear in next-generation checkpoint systems. One element that holds special promise is digital twin technology. A digital twin is essentially a digital representation of a physical asset. Using data taken from IoT scanners and devices, this technology can be used to create a command control center. For example, if the weather forecast calls for rain, the system user can click on “rain and haze,” and the simulated drivers on the display will slow down, providing a different outcome.

Service design for checkpoints

Achieving good service in a vast space like an immigration checkpoint may prove to be challenging, with tens of thousands of travelers and vehicles entering and exiting every day. Long queues, traffic jams and other inconveniences are to be expected. For border agencies, national security concerns will always overrule challenges delivering service levels or meeting the expectations of travelers.

Service design is one promising approach. In DXC’s service design for checkpoints, this approach follows five main steps:

  • Pre-work. To start, stakeholder reviews are conducted to understand the goals, mission and values of the service provider and agencies. Social mining is used to understand the expectations of service customers, and competitive analysis is applied to identify leading tools and practices.
  • Alignment and problem-framing. In this step, a big-picture overview is created for all the people, locations and associations that contribute to — or are impacted by — the service. In addition, an ecosystem map is developed to understand the relationships and interdependencies among customers and the business environment.
  • Discovery and mapping. The organisation gains a better perspective via tools that include service safaris, personas, empathy maps and customer journey maps. The end results include a problem statement and current-state service blueprint.
  • Ideation and mapping. This exercise starts with brainstorming as many ideas as possible. These ideas are then prioritised according to both their likely impact and the amount of effort likely to be required.
  • Evolution and piloting. A future-state blueprint is created to visualise potential changes in relationships among services and processes. It’s followed by idea validation, during which changes are analysed using simulation software.

Promising border technology

Thanks to a wide range of new and emerging technologies, intelligent border checkpoints will soon augment our travel life. Here are just a few of the exciting new technologies that could be used to modernise our existing checkpoints:

  • Artificial intelligence. AI provides smart analysis and predictions; it also reduces manpower requirements. AI is a broad field with a wide range of applications for checkpoints, including machine learning and neural networks. Possible use cases include:

– Video analytics with facial and iris scans provide fast and secure security checks. Cameras also can be used to analyse human behavior and facial features. The system picks up abnormal behavior or suspicious persons by analysing tell-tale signs. Then it can alert border officers to act before an incident occurs. In addition, AI can be used to automatically localise and detect suspicious items from X-ray images of the passengers and cargo. Finally, videos can be used to help detect road and infrastructure defects.

– Predictive analysis helps checkpoint officers predict cargo loads and the arrival times of buses and cars. AI-based optimisation models can be used too. They can be trained on historical loads of the border to determine the optimal amounts of resources and manpower needed to deploy, depending on the incoming flow of traffic and passengers. Security can benefit from anomaly detection, a subset of machine learning used to detect suspicious behaviors in normal routines, such as loitering, fraudulent paperwork and tampered vehicles.

– Reinforcement learning is a type of machine learning that enables agents to learn with interactive trial and error, using feedback from their own actions and experiences. In this way, the system can find a suitable action model that maximises the total cumulative reward of an agent. For example, agents can be shown the optimal routes to check.

  • Integrated IoT. This involves one or more networks of physical sensors and cameras to collect and share vital information about a checkpoint in real time. Information from these devices can be fed into either the digital twin or a centralised datamanagement platform for analysis and monitoring. The IoT network can also collect the vast amounts of data required for the success of AI, simulations and digital twin systems. Possible use cases for IoT devices include:

People/vehicle movement. IoT sensors can detect and track movement entering/exiting the checkpoint. This can be extended to monitor congestion and identify chokepoints in real time.

Record tracking. IoT sensors can automatically track paperwork and total time taken at each counter.

Environmental monitoring. Sensors can be used in a range of applications. This might include detecting air pollution, smoke or air temperature. It can also include more complex tasks, such as tracking the movements of birds on an airport runway. 

Integration with digital twin. A digital twin copies and virtualises an existing environment (such as machines, buildings and roads). Using data collected with IoT devices, accurate real-time information can be collected about a checkpoint. Then the entire environment can be displayed on a dashboard.  Human agents can use this dashboard both to improve the checkpoint’s processes and respond to problems quickly.

  • Next-generation kiosks. Unmanned and automated kiosks can be used for iris and facial recognition, as well as for passport validation. They may also feature cameras and card readers for paying road tax. Singapore’s Immigration and Checkpoints Authority (ICA) has implemented separate kiosks for motorbikes, and these can usually inspect an incoming motorbike without manual intervention. In the future, similar kiosks could also be used for cars, trucks and buses. In addition, scanners using both X-rays and gamma rays are being employed to quickly scan vehicles and their contents. Biometric scanners can help, too, by checking personal identities, clearances and even health. Passport control can also be automated.
  • Multipurpose booths and dynamic traffic management. With movable gantries and dynamic road signs, single-use booths can be repurposed to perform several different services. For example, a trucks-only booth could be transformed to also allow other vehicle types, such as cars and bikes, depending on the time of day. That might mean that a booth servicing trucks in the morning could also service cars in the afternoon and motorbikes in the evening. In each case, the correct traffic type would be redirected to the booth automatically, thanks to a combination of AI and reinforcement learning.
  • Drones and other unmanned aerial vehicles. These can be used to inspect vehicles and infrastructure, collect geospatial images, assess and surveil a site, or even send deterrent signals to stop potentially illegal activities. To prevent flocks of birds from impeding aircraft from taking off and landing, California Institute of Technology researchers in 2018 trained a drone to herd birds away from airports. Drones can also be equipped with sensors and attachments required to remotely perform the checks needed to clear a vehicle or passenger through the border. This removes the need for physical booths and personnel while increasing the number of concurrent checks that can be performed. 
  • Geospatial technology. The surge in geodata collection of information associated with a specific location — such as latitude and longitude — and the wide availability of such information mean that checkpoints can derive actionable insights from satellite images, cartographic maps, and data collected from IoT sensors and mobile phones. Geospatial technology has traditionally been used for city planning, agriculture and humanitarian aid. But it can also be used for managing and increasing security at border checkpoints, for example, by optimising IoT sensor deployment. Currently, the European Union’s Satellite Center (SatGen) and Frontex border and coast guard agency are collaborating to use geodata to detect and stop illicit trafficking and smuggling at border checkpoints across Europe.
  • Mobile apps. Smartphones and tablets can power both check-in apps and secure communications platforms for border officers. For example, Singapore’s ICA has implemented a digital arrival card that submits personal information, trip details and health declarations. Mobile apps also offer real-time information on traffic hotspots.
  • Digital Health passports. These are used to certify the fitness of a person for international travel. Stored data can include medical records, COVID-19 vaccination records and more. Blockchain technology can be used to store and access information. Both Singapore and Denmark have already announced that they will use blockchain to share their vaccination records with other countries.
  • Next-generation infrastructure. This is needed to support and integrate technologies including IoT, data management and 5G wireless networks. New infrastructures can also support clean and energy-efficient technologies, as well as smart traffic lights that help to clear traffic jams. An infrastructure enabled with 5G connectivity will be crucial, allowing for the implementation of tools and systems, such as drones and self-driving tow trucks, that can be remotely controlled in real time.


Border checkpoints today contain a long list of complexities and challenges, presenting a difficult balancing act between homeland security and operational effectiveness. Fortunately, checkpoints can take advantage of modern technology such as AI, IoT and drones. Looking ahead, these smart checkpoints will be able to handle heavy traffic while also maintaining high levels of security — and all with only minimal human intervention — to help border agencies strike the right balance between conflicting priorities. 

How DXC Technology can help

DXC is the preferred provider of leading-edge solutions for several checkpoint agencies worldwide. At DXC Innovation Labs, engineers constantly explore novel technologies, and develop prototypes and reference architectures to transform industries. Our innovative solutions are the key to navigating the new complexities that an increasingly globalised world brings. Partner with DXC Innovation Labs today, and join us on the journey to next-generation checkpoint solutions. 

Learn more about DXC analytics.

About the Authors

Nicholas Ang Kang Jing

Nicholas is passionate about applying technological solutions to solve real world problems. In his present role as a Data Science Engineer at DXC Technology Digital Innovation Labs, he dabbles with AI innovations and full-stack technologies to accelerate clients' digital transformation journeys.

Shu Xiao Hua

Xiao Hua is an experienced business consultant with DXC for over 15 years. Driven by passion, she takes pride in providing the best business solution as possible. As a new addition to the Digital Innovation Labs, she works as Scrum Master to help managing the team’s various projects.

Brandan Tan Chuan Kiat

Brandan Tan, data science engineer at DXC Technology Digital Innovation Labs in Singapore with experience in developing simulation, chatbot and web applications across various industries, with a specific focus in AI/ML and computer vision solutions. Brandan holds a degree in Physics and a MiniMasters in Business Analytics.

Ching Jia Chin

Ching Jia Chin is a Full stack developer working with the latest and newest technologies. Jia Chin received a Bachelors in Computer Science with specialisations in Artificial Intelligence and Cyber Security from NTU School of Computer Science and Engineering.

Tracy Goh Yi Fang

Tracy is a UI/UX designer who is focused on delivering a positive impact to people's lives. Every day, she is excited to unravel more about people's thoughts and behaviours in hopes to create meaningful experiences for users.