Introduction

Global trade relies heavily on containerised transport, yet moisture-related damage during shipment remains a costly and often underestimated risk. As sustainability targets tighten and data-driven decision-making becomes standard, the ability to predict and prevent such damage is more critical than ever.

This master thesis, conducted in collaboration between Absortech and Niftitech, explores how structured sensor data – including temperature, humidity, light exposure, and geolocation – collected from container shipments can be transformed into actionable insights. Through statistical analysis, classification techniques, and predictive modelling, the aim is to support smarter desiccant selection, optimise cargo protection, and contribute to reducing the environmental impact of global logistics. The project is ideally suited for two students but can also be carried out by one.

Background

Absortech collaborates with customers to perform shipment tests aimed at validating container dimensioning and proving the performance of desiccants in real-world conditions. These tests are essential for ensuring that goods are protected during transport, helping to avoid costly moisture-related damages such as mould, mildew, corrosion, and packaging deterioration. Throughout the shipment process – before, during, and after – detailed data is collected on cargo type, loading configuration, climate conditions, and the status of goods upon unloading. Climate data recorded inside the container includes:

• Temperature
• Relative humidity
• Light – to see container closing and opening
• 4G/5G positioning

This data is continuously gathered and used to support customers in selecting the correct desiccant configuration, thereby reducing waste and lowering the CO2 footprint of their shipments. ln addition to ongoing tests, there is a growing repository of historical data containing valuable insights and trends that the company aims to extract and analyse.

Scope

The scope of this thesis is to apply statistical methods, Al techniques, and/or predictive modelling to identify patterns and correlations within the dataset. The goal is to classify shipment profiles and develop models that forecast moisture risk based on known variables. This includes:

• Climatic classification of origin-destination routes (e.g. Hot-Cold, Cold-Hot)
• Trend analysis of condensation risk linked to seasonal variation, cargo type, and initial container climate
• Predictive modelling using historical sensor data to optimise desiccant selection and container setup

The thesis will also explore how physical shipment conditions can be integrated to enrich the dataset and improve model accuracy. Ultimately, the aim is to transform raw shipment data into structured insights that enable smarter, more sustainable logistics decisions.

Company Presentations

Absortech in Falkenberg

Absortech is a global leader in moisture damage prevention, dedicated to protecting cargo during shipping and storage. With a strong focus on innovation, the company goes beyond traditional desiccant solutions by offering data-driven strategies that reduce waste, cost, and environ mental impact across supply chains.

At the heart of Absortech's approach is the Peace of Moisture Mind® (POMM) methodology – a structured, analytical process developed over decades of industry experience. This method ensures that each solution is tailored to the specific needs of customers, combining analytics, expert knowledge, and high-performance products to eliminate moisture-related losses.

Arbortech's vision is to create sustainable and efficient supply chains by eliminating moisture damage and the waste it causes, ultimately contributing to a more responsible and resilient global logistics ecosystem.

Niftitech in Lund

Niftitech is a Swedish software company dedicated to creating intelligent, purpose-built digital solutions that simplify, enhance, and transform the way organizations work.

Founded in 2015 with deep roots in agriculture technology, Niftitech has evolved into a trusted development partner across industries – from precision farming to Al-driven innovation in the public sector.

At the core of Niftitech's philosophy is precision and adaptability. Every system is tailored to real user needs, built to be robust, easy to maintain, and intuitive to use. This customer-centric approach has earned the company a strong reputation for reliability, flexibility, and innovation.

Through collaborations with leading partners such as Al Sweden, Intel, Hushållningssällskapet, White Architects, Svensk Kolinlagring, and DataVäxt, Niftitech continues to push the boundaries of applied Al and data-driven technology. Projects like SveaGPT, the Al assistant for the public sector, and Markkartering.se, a precision agriculture platform used by thousands of farmers, exemplify how the company combines cutting-edge technology with real-world impact.

Niftitech's vision is to empower organizations with smart, sustainable software that frees up time, optimizes processes, and enables people to focus on what truly matters - creativity, collaboration, and progress.

Contact

If you are interested in this master's thesisi project, please contact Henrik Pålsson, Professor in Packaging Logistics via e-mail: henrik.palsson@plog.lth.se