This is a repository copy of Hydrogen production from bio-oil: a thermodynamic analysis of sorption-enhanced chemical looping steam reforming.

Presents research into the improvement of flow‐line assembly systems. Aims to understand and improve the design and control of manually intensive flow‐line assembly in the clothing industry. A simulation model of the progressive bundle system has been constructed, incorporating operator performance variations and learning effects, machine failure and repair, operator absenteeism, quality failure and supervisory control. While the operator performance data and the stochastic variables are handled satisfactorily within the simulation, the problems of control are not handled well by conventional discrete event modelling techniques. Adopts a knowledge‐based approach to control in which an online computerized supervisor exercises control over the execution of the simulation run. Complex system models are not easy to validate and a four‐stage approach is used to demonstrate conformance with real‐world systems: qualification; face validity; modular validation; and time‐series system behaviour. Discusses applications of the model and the results of experiments with a line starting work on a new style.

The problem of scheduling manufacturing systems where the performance, or indeed, capacity of a production resource is subject to stochastic change, is the subject of this paper. Typical of such resources are those which are dependent upon labour intensive processes.

Explains that the improvement of flowline assembly systems provides the context for this research: to understand and improve the design and control of manually‐intensive flowline assembly in the clothing industry. Constructs a simulation model of the progressive bundle system, incorporating operator performance variations and learning effects, machine failure and repair, operator absenteeism, quality failure and supervisory control. Notes that, while the operator performance data and the stochastic variables are handled satisfactorily within the simulation, the problems of control are not handled well by conventional discrete event modelling techniques. Adopts a knowledge‐based approach to control, in which an online computerized supervisor exercises control over the execution of the simulation run. As complex system models are not easy to validate, uses a four‐stage approach to demonstrate conformance with real‐world systems: qualification, face validity, modular validation and time‐series system behaviour. Discusses applications of the model and the results of experiments with a line starting work on a new style.

The paper discusses the design, implementation and validation of FLEAS, a flowline environment for automated supervision of clothing manufacturing systems. The paper argues for a mixed initiative approach to system control which incorporates both a scheduling component, based on local search, and a simulation component which dynamically tests the validity of the supervisory decisions made by the scheduler.