Computing is highly integrated into industrial processes, and the development of versatile and scalable industrial machinery relies on equally versatile and scalable software that efficiently utilizes the computing platform embedded in the machinery. Any conventional computer today, even if designed for industrial use, has a shared memory or multi-core parallel architecture. This architecture must be efficiently exploited in the design and development of industrial machines, since in addition to controlling common systems such as robots and conveyors, these systems need to collect and process information from the environment via increasingly complex sensors. These sensors provide ever greater amounts of information, and the processing is becoming more complex, which increases the overall computational cost. To meet the temporal requirements of such systems, this work presents a software design for an industrial machine equipped with at least one hyperspectral sensor. This industrial equipment is designed as an add-on that must be adapted to the needs of the industrial plant into which it is integrated.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Heterogeneous Thread-Based Parallel Software for Scalable Industrial Machinery

  • Miguel O. Martínez-Rach,
  • Otoniel López-Granado,
  • Carlos Pérez-Vidal,
  • Adrián Sarriás-Pérez,
  • Héctor Migallón

摘要

Computing is highly integrated into industrial processes, and the development of versatile and scalable industrial machinery relies on equally versatile and scalable software that efficiently utilizes the computing platform embedded in the machinery. Any conventional computer today, even if designed for industrial use, has a shared memory or multi-core parallel architecture. This architecture must be efficiently exploited in the design and development of industrial machines, since in addition to controlling common systems such as robots and conveyors, these systems need to collect and process information from the environment via increasingly complex sensors. These sensors provide ever greater amounts of information, and the processing is becoming more complex, which increases the overall computational cost. To meet the temporal requirements of such systems, this work presents a software design for an industrial machine equipped with at least one hyperspectral sensor. This industrial equipment is designed as an add-on that must be adapted to the needs of the industrial plant into which it is integrated.