PLANT

Wireless that works

Recent technology innovations overcome impediments to wireless networking in automotive plants.


To be competitive in today’s business environment, manufacturers require systems that quickly adapt to change but meet the increasingly stringent customer demand for high-quality, customizable, low-cost products that can be delivered quickly. However, the inherent complexity of manufacturing systems make it challenging to do so, but recent innovations in sensing technology show considerable promise of relief.

New devices contain multiple heterogeneous on-board sensors that are networked through wireless links and are deployable in large numbers. System-wide deployment of these devices is referred to as distributed sensing, and the whole infrastructure is called a distributed sensor system.

Wireless microsensor networking has been identified as one of the most important technologies of the 21st century. Indeed, it has resulted in a data-rich environment with both temporally and spatially dense information providing unprecedented opportunities for improvements to product quality and productivity in the automotive industry.

But successful applications of smart sensors and communication protocols may not be effective in the harsh, uncertain, dynamic environments typical of automotive plants. In military and defence sensor networks and in wireless environmental monitoring systems, sensors are usually deployed in open fields or in the ocean where systems are essentially free from interference. In manufacturing, various sources of noise and vibration, plus diverse electrical and electronic systems, greatly influence the performance and efficiency of wireless systems. Because conditions are so different, significant research is needed to develop wireless networks that will work. But it will take more than hardware to address the challenges posed by an automotive plant.

New control software is needed to improve the flexibility and responsiveness of the manufacturing systems.

Since 2008, the AUTO21 Network of Centres of Excellence has supported a distributing sensing and control project for automotive factory automation involving researchers from the University of Windsor, the University of Calgary, and the University of Western Ontario. Several private sector firms and public sector partners also provide support. The goal is to develop an efficient shop-floor sensing and control system to support real-time decision-making.

Project researchers have already developed several technologies to address implementation challenges. Their framework involves matching the control model more closely to the physical system of widely distributed devices in an environment that’s prone to disruptions. With this model, control is achieved by the emergent behaviour of many simple, autonomous and co-operative agents that initiate their own activity. Fitting multi-agent systems technology to manufacturing problems has led to many applications in this domain.

A node technology has been developed that’s capable of locating mobile objects in 3D co-ordinate space in real-time, indoor environments. In addition to location tracking, the technology also has data networking capability and is simpler and relatively more cost-efficient than existing solutions.

Researchers are also working on a new class of intelligent devices and software modules that support plug-and-participate functionality with wireless connectivity. This involves the development of agent hosts based on Java and Jini technologies that plan and execute their actions and cooperate to achieve increased productivity.

Although Jini was originally developed by Sun Microsystems to support self-configuration within office computer networks, the project team focuses on its use in the resource-constrained, real-time processors of industrial automation systems. A methodology for intelligent wireless connectivity is based on self-organizing 802.11 networks integrated with the Java Applications Development Environment (JADE) platform.

Efforts are also underway to develop miniaturized smart camera technology for use in quality control, assembly and robotic guidance.

Collectively, these innovations in technology are overcoming the impediments to establishing distributed sensing systems in automotive plants.

Jonathan Wu is a professor at the University of Windsor and leads the AUTO21 project “Distributed Sensing and Control for Automotive Factory Automation.” Visit www.auto21.ca.