IPSO CHALLENGE 2013 Interviews

Smart Factory_CH2013FAST Solutions for Smart Factories

Contestants: Dr. Jose Martinez Lastra, Gerardo Santillán, Luis Gonzalez, Tomas Sedlacek
Affiliations: FAST Laboratory at Tampere University of Technology
Semi-Finalist

 

Summary

FAST Solutions for Smart Factories is the result of the work done at FAST Laboratory towards a fully IPv6 manufacturing floor. We proposed a set of solutions that take the IPv6 protocol to embedded devices for industrial monitoring and control, data acquisition, data processing, and enterprise integration of industrial assets. FAST has focused on developing solutions that add connectivity to production in the factory for energy-efficient manufacturing systems.

SmartFactory_001

A big part of the groundwork that gives foundations to the potential of the IoT is being laid to the manufacturing industries. The IoT has enabled the innovation in many fields of the manufacturing levels. At FAST Laboratory, we believe that the implementation of “Smart Factories”, capable of self recovery, goes along with the deployment of IP based technologies, for this reason we have developed a number of solutions based on the idea of having everything connected to a network so that the information from “things” can be combined with intelligent systems in order to create knowledge.

We proposed a set of solutions that take the IPv6 protocol to embedded devices for industrial monitoring and control, data acquisition, data processing, and enterprise integration of industrial assets. We have focused on developing solutions that add connectivity to production in the factory for energy-efficient manufacturing systems. We have developed a fully IPv6 production line where devices and applications are networked using standardized and open protocols.

By bringing IPv6 smart objects in manufacturing, our solutions are able to take advantage of powerful features such as the web-services, that allows horizontal and vertical integration of automation systems with common IP-based standards and technologies, and self-discovery and self configuration that, as a result of W3C standards, have made possible automated qualitative and quantitative improvements in the overall functionality of manufacturing systems.

A big part of our solutions are 6LoWPAN based, The IPv6 over Low Power Wireless Area Networks (6LoWPAN) is an interesting standard that brings IP directly down to small wireless devices. Knowing that there are not enough of the current IP-format addresses to extend to the IoT, 6LoWPAN brings the IP version 6 to give an address to every device. The wireless communication technology that 6LoWPAN utilizes is the IEEE 802.15.4, a promising standard for the lower (physical and link) layers.  In general, it can be said that 6LoWPAN is the transmission of IPv6 packets over the 802.15.4 wireless standard.

Before the 6LoWPAN-enabled WSN´s capabilities can be exploited, significant problems need to be solved, including simple adaptation and integration issues as well as the creation of mechanisms that enhance the management of resources at the lower layers of the 6LoWPAN protocol stack. And it is at this low layers in which one of our solutions focuses: an adaptable wireless packet scheduler designed specifically the different traffic conditions of a smart factory.

The use of IPv6-based wireless sensor nodes attached to moveable objects on the factory floor results in new applications for the real-time monitoring of assets. The optimization of production lines has been always one of the main priorities of the manufacturing industries. Nowadays it is possible to identify faults and incongruences at the production lines like never before. Having numerous pallets with products on top of them represents for each vendor considerable assets and thus their continuous monitoring to prevent damages or losses become one of their main objectives.  Unfortunately, traditional monitoring based on a network of distributed sensor cannot simply provide desired results. There is actually no information about pallet movement or position in between those sensors. If certain pallet stacks somewhere, falls down from conveyor or it is in the worst case even taken out by irresponsible operator, the conventional monitoring system would never be capable to detect it and react appropriately. Therefore, a problem of continuous real-time monitoring of pallets became a main motivation for this design at FAST.

In order to detect pallet position while it moves throughout the conveyor system, certain sensors must be installed to provide desired information. Since the pallet is still in motion there are only two reasonable ways of its detection. There could be either machine vision applied to track the pallet continuously or each pallet can be equipped directly by small wireless sensors. While first solution would probably provide very nice results, it is very expensive, for it requires the installation of high cost cameras meanwhile the latter solution is significantly cheaper, since the prices of MEMS sensors are very low nowadays and does not require any special hardware installation. Our pallet positioning system estimates the position helped with the information of two sensors: 6LoWPAN accelerometer/gyroscope and RFID readers connected to an IPv6-based industrial controller.

Our solutions use the IP in order to be networked. One of the main benefits of using IP is the possibility to bring the advantages of networking down to small devices such as our WSN.

Our WSN has connectivity and communication with other networked devices: it is possible to access both, our sensors (through their IPv6 addresses) and their interfaces (all the applications are accessible through internet) using computers, smart phones or tablets. This allows having a constant monitoring of the factory floor from any part of the world. This also means that we are exploiting the internet access sharing feature brought to our solution by the usage of the IP.

IP data sharing capabilities changes the factory floor monitoring, making the plant being more monitored and controllable: with our solutions is possible to know the position of a pallet in a factory floor in Finland from anywhere. It is also possible to have a data security and a data management, the access to the information is controlled by an administrator and it is possible to control what information can be shown to whom.

IP brings the benefit of hardware sharing: multiple nodes using the same edge router.
Interoperability is one of the problems addressed by using IP. Our design is based on the integration of different technologies that converge into the usage of the IP. The internet convergence brings a huge amount of possibilities.

We try to give a greater view of the plants behaviour so that it is possible to have a better control of assets and a better monitoring, translating this into more efficient production processes.

Our solutions are aimed to be used by factories that look for the highest efficiency. We are
proposing solutions that bring the Internet of Things to the manufacturing floor in a way that any other products had, we believe that by implementing this solutions today we are building the smart factory of tomorrow.

Press: http://www.noodls.com/view/7E88223F5D15F104202D9B74B3B7F259BB816E5D