Simulation tools for understanding logistics and create explanations for logistics operations are often developed under a state of stability, equilibrium and linearity whereas logistics are considered non-linear and complex. Agent based modelling provides the opportunity to build rigorous computer simulations of non-equilibrium systems and unlike other simulation methods, no agent has full knowledge about the system it operates in and the agents act on heuristics rather than on rationality, much like customers and firms in the real world. In the last twenty years the development in knitting technology has made it possible to produce entire garments in one piece with no need for time-consuming cutting and sewing processes afterwards. This gives new opportunities to produce knitted garments closer to the customer and hence being able to respond faster to shifts in demand for knitwear. Agent Based Modelling will provide an excellent opportunity to simulate a demand chain of the kind where the customer designs the garment and the garment is produced instantly. The aim is to create and present an agent-based model for the logistics operations in the Knit On Demand research project. The simulation will not only help to gain insight about the dynamics and emergent behaviours in a textile demand chain but also in demand chain management as a whole, as well as providing a base to test theoretical frameworks.