V.I. Vitanov

Survey: Allocation of quality control stations in multistage manufacturing systems

The allocation of quality control stations (AQCS) in multistage manufacturing systems has been studied extensively over the decades. This paper reviews the existing approaches, models comparison and solution techniques applied in AQCS. The relevance of the models and the effectiveness of the inspection strategies are examined by developing a generalised model. The conducting simulation experiments show that as the number of workstation increases the processing time to solve the problem increases significantly. This led to the development of a heuristic algorithm with local search. The performance the heuristic was compared with the optimization method based on complete enumeration method (CEM). It was found that the heuristic method can derive an acceptable solution significantly faster than the CEM. The review has shown that the most common techniques used are dynamic programming and non-linear programming. The paper suggests some biologically inspired optimisation algorithms can be of interest for further study.

Emergency response networks for disaster monitoring and detection from space

Numerous man-made and natural disasters have stricken mankind since the beginning of the new millennium. The scale and impact of such disasters often prevent the collection of sufficient data for an objective assessment and coordination of timely rescue and relief missions on the ground. As a potential solution to this problem, in recent years constellations of Earth observation small satellites and in particular micro-satellites (<100 kg) in low Earth orbit have emerged as an efficient platform for reliable disaster monitoring. The main task of the Earth observation satellites is to capture images of the Earth surface using various techniques. For a large number of applications the resulting delay between image capture and delivery is not acceptable, in particular for rapid response remote sensing aiming at disaster monitoring and detection. In such cases almost instantaneous data availability is a strict requirement to enable an assessment of the situation and instigate an adequate response. Examples include earthquakes, volcanic eruptions, flooding, forest fires and oil spills. The proposed solution to this issue are low-cost networked distributed satellite systems in low Earth orbit capable of connecting to terrestrial networks and geostationary Earth orbit spacecraft in real time. This paper discusses enabling technologies for rapid response disaster monitoring and detection from space such as very small satellite design, intersatellite communication, intelligent on-board processing, distributed computing and bio-inspired routing techniques.

Buffer capacity allocation using ant colony optimisation algorithm

This paper presents an algorithm for the near-optimal allocation of buffer space to an assembly line by means of the ant colony optimisation (ACO) paradigm. Uniquely, the algorithm has been designed to work in conjunction with a simulation model and is adapted to have both combinatorial and stochastic problem-solving capability. The simulation model was developed using the WITNESS simulation package and serves the purpose of an objective function evaluator, encapsulating the dynamics of the line and enabling the production rate for a particular buffer configuration to be determined. Finally, the WITNESS Optimiser module was used as a benchmark in validating the ACO algorithms performance. In the simulation experiments conducted, ACO attained slightly better results overall.