Web & mobile apps

to

automate

the passenger

boarding process

33%

faster baggage weigh-in

40%

cut in queuing time

18%

increased passenger satisfaction
Web Application
Native Android development

About 
the client

The founders of the startup aimed to help airlines automate the passenger boarding process at airports.

Goals

Custom Software Development
Improve efficiency in the baggage weigh-in process to reduce queues and delayed departures
Custom Software Development
Automate daily tasks for cabin crew and airline managers
Custom Software Development
Minimize human errors in calculating additional airline services
Custom Software Development
Enhance passenger satisfaction and overall operational efficiency for airlines

Needs

Develop a multi-platform AI based application for airport employees that will be user-friendly and speed up the check-in and baggage loading process.
Solution Development

Proof of Concept Development

Using the MaskRCNN architecture and Detectron2 AI model, we identified luggage boundaries and calculated sizes relative to a standard reference sticker.
To enhance accuracy, we adjusted for the angle distortion in photos, reducing the error by 5%, achieving a level of precision that meets operational requirements.

“In the PoC phase, we aimed to accurately measure passenger luggage size from photos with a maximum error of 10%.”

Andrew Laminski, CTO at Gilzor.

UI/UX Design

The idea of the startup was to build an Android application for cabin crew and a web admin panel for airline operators.

The mobile application

is designed to provide users with limited information and functions that can be used during boarding.

The web application design

is adapted to all commonly used browsers and screen resolutions. Web design contains a lot of data, tables and spreadsheets that have been designed in such a way that they can be easily compiled and modified when necessary.
Solution Development

The app functionality

Display additional services calculation

The mobile app provides calculations of extra fees to avoid human errors & financial losses.

MAKE A PHOTO

with the help of users’ phone cameras.

Display baggage information

As a photo of baggage is done, the mobile app shows whether the baggage is within limits or overweight.
Solution Development

Mobile Development Tech Stack

Language

Kotlin

Frameworks

JetPack Compose
Room
LiveData
Coroutines Flow
Hilt

Assistive Technology

CI/CD
Feature Flags
Solution Development

Web application development

Web admin panel for Airlines Operators includes

User management

Adding and removing users as necessary.

Baggage limits

Setting specific weight and size limits for cabin baggage to ensure compliance with airline policies.

Additional services configuration

Setting different types and parameters for calculating additional services such as baggage wrapping, delivery, etc.

Overweight baggage management

Determination of baggage overweight cases to address them accordingly.

Statistics and payment tracking

Controlling the number of excess baggage cases, online payments, and other additional costs.

Hierarchical access structure

Creating appropriate access levels for different airline staff members.

Multitenancy

Support of multiple airlines with separate data and configurations to accommodate different clients.
Solution Development

Web application development Tech Stack

Languages

TypeScript
CSS (for Tailwind CSS)

Frameworks

React
Tailwind
CSS
Redux
Solution Development

Backend development

The backend of the app is responsible for processing the image recognition data, calculating baggage weight, and managing the admin functionalities.

Users role system

is designed with granular access and effectively limits access to system resources.

Determining luggage size

from a photo using OCR AI technologies and an acceptable accuracy of less than 10%.

Calculation of additional services

for passengers and displaying calculation results to airport employees.

Integrated AWS CloudWatch

monitors system metrics to resolve potential errors within minutes.

Payment processing system

uses third-party services, processes user requests and notifies in case of errors.
Solution Development

Backend Development Tech Stack

Languages

Node.js
TypeScript
Python

Frameworks

Nest.js
FastAPI

Assistive Technology

OpenCV & Deep Neural Network
MySQL
Amazon S3
Elastic Container Service
Docker
Solution Development

Quality Assurance

We tested the product manually to ensure the application's accuracy and reliability.

The QA process covered approximately

15% of the project duration

Postman
Jmetr
Figma
Dew Tools
Swagger
Xmide
Solution Development

Time & Human Resources

Fulltime

support

Achieved

results

33%

faster baggage weigh-in process

40%

less time queuing

18%

higher passenger satisfaction
Decreased number of delayed departures

43%

less human errors in calculating additional services