Flavors Technology wins NIST ATP award

Citation

(1998), "Flavors Technology wins NIST ATP award", Assembly Automation, Vol. 18 No. 1. https://doi.org/10.1108/aa.1998.03318aab.013

Flavors Technology wins NIST ATP award

Flavors Technology wins NIST ATP award

Flavors Technology Incorporated. Manchester, New Hampshire, USA, has received an award from the National Institute of Standards and Technology (NIST) to develop a manufacturing agility server. The award comes from the autumn Motor Vehicle Manufacturing Technology (MVMT) competition of the Advanced Technology Program (ATP).

As a non-regulatory agency of the US Department of Commerce's Technology Administration, NIST promotes economic growth by working with industry to develop and apply technology, measurements and standards. Flavors proposes to build a computational system using autonomous software agents to dynamically reassign operation from failed manufacturing devices in automotive assembly lines. For example, the failure of a single robot in an automated manufacturing cell can force an entire production line to stop, even though other robots on the line could fill in for the failed device. A software system that could dynamically reassign operations along a production line would lower the cost of automotive body assembly and have an impact on other types of manufacturing as well.

Flavors Technology and its subcontractors propose to develop a systems architecture base on a Manufacturing Agility Server (MAS) and autonomous software agents that would permit manufacturing operations to be reassigned dynamically to similar manufacturing equipment. MAS would receive status information from the network of device controllers within a manufacturing cell or along a production line. If a device fails or is unavailable because of maintenance, MAS would reassign the task.

Conventional approaches to reconfiguring manufacturing processes, such as classic optimization schemes, are not used in practice owing to their extensive calculation requirements and extreme sensitivity to process dynamics. In contrast. MAS would rely on the interaction of independent software agents ­ software agents that are a spinoff of artificial intelligence research ­ that can respond to events independently. Such a system could also support dynamically reconfigurable production lines. Once the core MAS technology is developed, it will be tested in a simulated production environment, and ultimately will be integrated with the control systems of an automotive assembly line.

Within the USA MAS technology could significantly increase welding line productivity, reduce vehicle time-to-market. and improve vehicle quality. The MAS Project will run for three years. The estimated cost of the project is $2,522,000, with $1,988,000 coming from ATP funds.

For further information contact: Doug Currie or Bob DeSimone. Tel: +1 (603) 647 1270; Fax: +1 (603) 647-6180; e-mail: e@flavors.com or crd@flavors.com