Research Overview

Building intelligence for logistics

We build AI that reads operations and delivers results, moving towards autonomous execution.

What our research is building

Understand the operation
Models that read across scorecards, workforce, fleet and cost data to see what is driving performance and where attention should go.
Move towards autonomy
From recommending actions to executing and coordinating work, through software agents, autonomous workflows and, over time, physical systems.
Learn from every outcome
Each decision, intervention and outcome is feedback. Our research explores how models learn what works, where and under which conditions.

What happened?
Building a clear picture of the operation from fragmented data.
Why did it happen?
Telling symptoms apart from the causes underneath them.
What should happen next?
Finding the interventions most likely to improve the operation.
What worked?
Connecting each decision to its outcome, so the next one is better.

Research programmes:

Areas of study:

Four areas of research, one aim: AI that understands logistics operations, reasons across real-world constraints and increasingly acts within them.

How AI can understand an operation from fragmented data, find what is driving performance and decide where attention should go.

How AI can move from recommending actions to executing and coordinating work across logistics operations.

How models can learn from decisions, interventions and outcomes, and improve with every operation they meet.

How AI in the physical world stays dependable, measurable and aware of its limits as it takes on more.