Big data, along with artificial intelligence solutions is expected to play an important role in the fourth industrial revolution. But as much appealing as this sounds, there are certain
challenges in the collection, integration, and sharing of big data in this field.
Speaking to
a&s, Philipp Unterhalt, MD of
HAHN Group and Interim CEO of
Rethink Robotics, pointed out that connectivity and transformation into a usable format are key when it comes to an integrator such as HAHN that uses a vast variety of different components.
"Striving towards predictive analytics however is requiring more than that,” Unterhalt said. “We need to rethink the engineering process hand in hand with the business model. Take a temperature sensor as an example. Currently, it might be sufficient to have only the information available, whether the temperature is within the defined threshold. The output is yes or no. For predictive analytics, it might be important to understand what the specific temperature is, and how it is influencing the production process, the yield. In such case, the output is specific root cause analytics of the reason why the yield was affected all the way to suggestions before it happened (predictive).”
Complications of legacy systems
Other industry professionals said some of the major challenges are related to retrofitting devices into existing systems in a cost-efficient manner.
Heiner Lang, Senior VP,
Business Unit Automation
and Electrification Solutions,
Bosch Rexroth
“There are several challenges here,” said Heiner Lang, Senior VP of Business Unit Automation and Electrification Solutions at
Bosch Rexroth. “Greenfield approaches — where manufacturers build completely new factories — will be the exception. In most cases, manufacturers will continue to use already installed equipment, which will have to be integrated into these connected environments in an economic manner. Solutions like Rexroth IoT Gateway software, in combination with sensors, establish real-time monitoring of process data such as temperature, pressure, vibration, etc., of the machines. The brownfield approach enables manufacturers to reduce the complexity and introduce Industry 4.0 gradually.”
Big data only create added value in combination with the domain knowledge of automation and processes to turn these data into information, as well as to ultimately draw up hands-on conclusions with regards to optimization.
Sunil Mehta, GM of Automotive Business Development, Factory Automation and Industrial Division at
Mitsubishi Electric India, elaborated on this further, adding that in most of the manufacturing industries there are as of now several old types of equipment being used on the shop floor and the systems, with controllers used in these old machines not able to communicate with the outside world.
“In brownfield projects, there is a mix of old and new equipment in production,” Mehta said. “In case of greenfield projects, it is possible to decide upon controllers, network protocols, etc. so that seamless integration of all the equipment will be possible. In many of the manufacturing industries today there are different kinds of controllers used and these controllers support different kinds of protocols. In this case, it becomes difficult to communicate to upper application layers such as SCADA (Supervisory Control and Data Acquisition Systems), MES (Manufacturing Execution System), ERP Systems, etc.”
Quite often, changing old controllers, dedicated microcontrollers and old PLC becomes important. They will need to be upgraded so that equipment can be integrated with each other and then to the upper application layer.
“New technologies needs to be embraced to achieve Industry 4.0 and to collect big data,” Mehta added. “Industrial Ethernet is becoming more popular and available for carrying big data in the manufacturing process and this technology needs to be adopted. One important step is to form a cross-functional team where manufacturing, maintenance, project engineering and IT members need to work together for a common goal. We must have robust and uninterrupted internet services and networks to implement Industry 4.0 in manufacturing industries.”
Something else that is a requisite for successful implementation of Industry 4.0 is a certain kind of discipline from all operators on the shop floor. The operators and supervisors should work together to equip themselves with the right skills to handle such systems. The business model needs to be implemented in such a way that it is based upon real-time data generated by the system.