AI in travel: easier travel document processing with a no-code AI tool

Travel volume is at an estimated one-third of pre-pandemic levels, with over 50 million travellers a month this year. Multiply it by the cost of labour to manually process the growing number of travel documents, like PCR tests and PLF forms, and the total costs are staggering. Can AI make it better?

Countries around the world are experiencing a resurgence of cross-border travel. The UNWTO World Tourism Barometer estimates 54 million international travellers in July of 2021. While it’s 67% lower than the same month in 2019 before the pandemic hit, things are slightly different this time. New precautions are in place as the world continues to tide through a post-pandemic new normal. And the cost to maintain new operations are sky high.

 

Our global post-pandemic travel resurgence is taking a different tack

Complex sets of measures are taken by different aviation authorities to keep the spread of Covid-19 in check. Almost all passengers are required to take various types of Covid-19 tests prior to arrival and departure, such as polymerase chain reaction (PCR) tests. Test results or certificates must then be submitted or uploaded to the relevant authorities for verification.

At the same time, passengers also need to declare information surrounding their travel plans. These include travel and accommodation information, proof of vaccination and testing reference numbers—at the granular level. Passenger Locator Forms are what capture this data. Authorities rely on these to facilitate contact tracing efforts upon risk of exposure.

Dubai PCR travel document

Polymerase chain reaction test report

GR PLF travel document

Passenger Locator Form

More travel documents means more document processing

For public health, immigration and aviation authorities, the step up in regulations means more paperwork. Paperwork that needs to be scanned, verified, and transferred to third-party systems manually. This begets the obvious question, why isn’t the process done online?

A large bulk of these checks remain document-based by nature. In either paper or digital formats, these documents provide a universal way to collect information. Documents accommodate all travellers and service providers who need not rely on internet connection or power. They’re easy to fill out or submit anywhere, by anyone, which are particularly important considerations for the aviation industry. Because of these factors, documents provide a layer of security as compared to e-forms or digital portals. In turn, documents need to be manually processed, costing valuable overheads.

AI can process travel documents easier than you think

Automated document processing tools, also known as Intelligent Document Processing, use AI to automatically extract data from travel documents and more.

Intelligent document processing (IDP) consists of a host of business solutions that process data from all types of documents by capturing, organising and extracting the data therein using a range of AI technologies.

Enterprise Grade IDP Solution TAIGER

Enterprise-grade intelligent document processing workflow

So, it takes a Covid-19 PCR document certificate for example, and turns it into raw, usable data as illustrated below.

travel document processing

Data is automatically extracted from PCR test result document for further use

A no code approach makes it even easier

IDP solutions are commonly used across industries and verticals, but often require costly investment in IT talents. An easier way to access IDP and its benefit is to place it in the hands of anyone, such as with Omnitive IDP.

Omnitive IDP is hosted entirely on no-code web interface to conduct travel document processing. No IT engineers or data scientists needed, no complex code to break down, and no technical interface to navigate.

Users can also build extraction models on a simple interface, entirely without code. Virtually anyone, from public health employees to immigration officers can build and use models on demand, according to their needs and schedules.

Omnitive IDP interface for travel documents

Omnitive IDP training interface

IDP solutions are commonly used across industries and verticals, but often require costly investment in IT talents. An easier way to access IDP and its benefit is to place it in the hands of anyone, such as with Omnitive IDP.

Omnitive IDP is hosted entirely on no-code web interface to conduct travel document processing. No IT engineers or data scientists needed, no complex code to break down, and no technical interface to navigate.

Users can also build extraction models on a simple interface, entirely without code. Virtually anyone, from public health employees to immigration officers can build and use models on demand, according to their needs and schedules.

Three main functionalities accelerate travel document processing

On Omnitive IDP’s, users can do three key tasks to automate travel document processing at speed.

1. Self-train extraction models with minimal effort, expertise, or resources
2. Automatically extract data from travel documents to derive usable data
3. Validate extraction accuracy and detect anomalies without any manual processing

To learn more about each functionality, download our travel document processing ebook here for a deeper dive.

Unlock travel document processing

See Omnitive IDP solutions
for the travel industry
‘Hybrid AI’ is what makes this no-code approach possible

Making this no code approach possible ties back to how Omnitive IDP is built. Most industry IDP solutions predominantly use Machine Learning, a subset of AI that is statistical by nature. On the flipside of Machine Learning and other statistical AI technologies is symbolic AI.

Statistical Machine Learning approaches ‘learn’ based on samples because they’re trained using data to find patterns. On the other hand, symbolic AI uses pre-trained rules, or logic, to understand data. Symbolic AI understands the rules of language like a human brain, to effectively read data such as text.

TAIGER’s IDP tool adopts this unique ‘hybrid AI’ approach. Omnitive IDP specifically combines Machine Learning with symbolic AI technologies like advanced natural and semantic language processing algorithms. These different AI technologies are used together to automatically identify, extract, clean, validate and store key information from unstructured and semi-structured travel documents.

Hybrid AI helps to make IDP accessible to anyone. It reduces the amount of document samples required to train extraction models. In turn, it reduces the time spent to develop ready-to-use models to a matter of hours to days.

The ‘hybrid AI’ approach also solves the main challenge of variability

As there are various types of Covid-19 tests, there are also multiple ways to get verification for having done one. Each country has their own standards for test and vaccination certificates provided by locally authorised clinics and providers.

Documents can take a different style or format to express a diverse range of data—i.e. high variability. As such, there’s no international standardisation, or even that on a nation-wide level for large countries.

Different types of PCR forms

Different formats of PCR test result documents

High variability spells trouble for typical Machine Learning-based IDP tools. Machine Learning-based IDP models learn patterns from large volumes of data (or documents samples in this context) within a very narrow scope.

Think tens to hundreds of thousands of document samples to read vaccination certificates issued by the Singapore government. The model would function superbly for just Singaporean certificates, but dismally for any certificates from all other countries.

Training and developing different models for each type of document is not feasible for the travel industry either. Each model could take months to build, considering the resources required. These all add up to massive costs that force companies to rely on the cheaper, manual alternative. IT talents are also highly involved in the process of model development and beyond, because these models must also be continually maintained when documents change over time.

 

Hybrid AI no-code IDP versus standard IDP solutions

Hybrid AI tackles the key challenges of IDP on all fronts. They’re far easier to build, without needing much resources in terms of document samples or specialised talent. Moreover, processing accuracy stays high above industry standards to further alleviate manual effort.

Omnitive IDP
Standard IDP solutions
Training samples required
100’s or less
10,000’s or more
Processing accuracy
Over 85%
80% or less
Speed
Few hours to a few days
Months

The result? Models highly versatile to automate travel document processing for a wide variety of travel documents. From the business point of view, it means less cost and much more scalability than standard automation solutions.

Get started with no-code travel document processing

The travel industry is only expected to go up from here as countries around the world find ways to live with the virus. As more countries open up and more precautions are set, organisations related to travel will see increasing volumes of data being generated and submitted, that need to be managed effectively.

Immigration and aviation authorities, public health organisations and healthcare institutions like hospitals have critical roles to play in our new travel resurgence. Travel document processing can be done easier, leaving us for more time for more meaningful tasks.

Unlock travel document processing

See Omnitive IDP solutions
for the travel industry

About TAIGER’s Omnitive IDP

Omnitive IDP is TAIGER’s next generation intelligent document processing (IDP) tool designed for anyone to build and use extraction models for any kind of document. What Omnitive IDP brings to the table is a a no-code approach to IDP that's faster and more scalable than any enterprise-grade IDP tool in the market.

Speak with one of our solutions managers today on how Omnitive IDP can be applied in your organisation today.

Sorry

Your browser is unable to display our site correctly.
It is best viewed on a modern browser, such as Chrome, Safari or Firefox.