Bid to enhance AI adoption in logistics

THE COMPETITION attracted over 1000 applications awarding only a handful of winners in logistics. Logidot’s winning project – in collaboration with the University of Bath’s centre for Smart Warehousing and Logistics systems, and a leading contract logistics company – aims to validate and develop an AI agent to support competency and AI adoption for various roles in the logistics sector.

It seeks to develop a customisable and scalable AI agent/virtual assistant based learning tool.

While large language models (LLMs) have enabled significant progress in natural language interaction, they remain prone to hallucinations — confidently producing inaccurate information. Overcoming these limitations is critical not just for safe deployment, but also for enabling future AI systems capable of continuous, context-aware collaboration with humans. Multimodal AI, grounded in real-time sensory data and guided by human oversight, offers a more reliable path. This “Physical AI” approach, exemplified by Logidot, fuses data from location sensors, operator-held devices, and human-in-the-loop feedback to deliver a location-aware assistant tailored to both frontline workers and managers — a first in the warehousing sector, says the firm.

Industrial players report major barriers to AI adoption – including skills and training challenges, lack of company-wide strategy, privacy and security concerns, information accuracy and overload and speed of integration with legacy systems and running costs.

To address these challenges, Logidot and the University of Bath are developing an AI tutor to support continuous upskilling and reskilling, particularly for seasonal and frontline workers. Using natural language and voice interfaces, the system offers fast, intuitive support for staff engaged in physically demanding roles. The AI assistant integrates a client-specific knowledge base with Logidot’s IoT platform — providing real-time visibility of assets and workforce location — and connects with an AI-powered warehouse simulator and execution system to enable seamless collaboration between human and automated operations.

This will enable:

• A personalised AI tutor for continuous upskilling and reskilling.
• Real-time task guidance and decision support tailored to user roles and locations.
• Enhanced managerial oversight, enabling supervisors to monitor team competency, training progress, and task execution through explainable insights.
• Intelligent system integrations, connecting frontline staff, warehouse management systems, and automation platforms to reduce errors and enable real-time coordination.

Logidot founder Dr. Niccolo Corsini, says: “Physical AI promises to be a key enabler and we are honoured to have been recognised as an innovative solution to help companies on this digital transformation journey.”

University of Bath’s Centre for Smart Warehousing and Logistics systems director Vaggelis Giannikas, adds: “AI, like many other technologies, can be extremely beneficial. Yet, we see time and again traditional industries to be lagging behind when it comes to its adoption. This project will help use our expertise in AI tools for industrial operations but also in technology adoption barriers to shape the future of smart logistics and warehousing.”