Live Data Is Improving Efficiency and Productivity on the Shop Floor, but There Are Challenges

Sensors that automatically collect and report data are being attached to systems at nearly every step of the manufacturing process and within support systems throughout the factory today. Production machines, forklifts, HVAC systems, conveyors—any piece of equipment with such devices as heaters, chillers, pumps, pressure vessels or throughput sensors—are able to monitor and communicate temperature, throughput, flow rates, pressures, energy consumption and other critical parameters, and in many situations address issues automatically, and much more quickly and accurately than ever before.

These automated and connected devices can be found in some form in most industrial facilities today ranging from automotive assembly to plastics injection molding to textile fiber spinning factories. Once these devices become connected, the emphasis here is on the word ‘automated.’ You won’t see anyone monitoring many of these devices manually. Rather, computers maintain everything in the background and take process adjustment action within the system almost real-time taking into account many elements of inter-related process data.

For example, in a sophisticated fiber spinning factory, thousands of spinning heads might be equipped with a variety of sensors that all feed back to a single process management center in order to allow related process variables to be controlled in a coordinated way. Recordings will show what process parameters need to be adjusted, such as when temperature is getting below process standards, the computers will make process adjustments without human intervention.

THE FUTURE OF INTERNET CONNECTED DEVICES FOR MANUFACTURERS
While sensor technology has been around for years, it is the rapid growth of internet connected devices driven by the expansion of wireless networks, both Wi-Fi and cellular, with their increasing coverage areas and data transfer speeds, that is allowing efficient inter-connection of process sensors throughout the factory, and equally important, in between inter-related sites that may be located far apart.  In addition, advances in the reduction in size of sensors is allowing them to be put into smaller spaces without having to do significant redesign, and allowing cost effective retro-fitting of equipment and facilities.

Costs are also going down. Ten years ago, if you bought a Wi-Fi card for a PC, you would have paid $100. Now you can almost get one free at Staples. The cost of sensor has also dropped dramatically. For example, some motion sensors now cost under a dollar and this cost is dropping 2% a year. This will greatly increase their use.

Also, communication and data standards are rapidly being developed. Agencies around the world are coordinating to standardizing data formats and communication protocols and manufacturers are adopting the standards to ensure that systems can be created in a flexible and efficient way, and improving the ability of MES and ERP systems to seamlessly use newly available data.

The ubiquity of these data systems allows for connectivity far beyond the factory.  Increasingly, systems automatically monitor performance of products once they’re sold and deployed in the field, allowing design and manufacturing improvements much more quickly. Think about when you buy an inkjet printer. As you’re loading the software after purchase, you may be prompted with a question: “Is it OK if we report usage data to the manufacturer?” Most people check “yes.” Along the way, if it frequently jams, this information may be going back to the manufacturer. So even without having to do a big product survey or checking to see how many printers are returned or how many consumer complaints there are, the manufacturer is collecting real-time data on uptime and the efficiency of their product and are able to make design or manufacturing process changes.

In the warehouse, there is data being collected on where the forklift is in the factory, how many hours a day it’s running, how many miles are being put on it, etc. Then, much like you would get an email from your dealership when your tires need changing, forklift distributors are working with customers to improve up-time by doing preventative/predictive maintenance based on usage data.  So on a Tuesday morning, a manager gets a call from the maintenance company saying it’s time for your forklift tune-up, or “there’s a reading on forklift #16 and we need to come out and take a look at it.”  Warehouse environments also routinely integrate RFID technology with these other technologies to improve materials and equipment tracking.

When Malaysian Airlines flight 370 mysteriously disappeared on March 8, its Rolls Royce engines had been sending performance data wirelessly back to Rolls Royce. This data ended up being used for very different purposes than initially intended, but this is illustrative of how, transparent to the user, products are conveying performance information back to the manufacturer so that efficiency and performance can be continuously improved.

CHALLENGES GOING FORWARD
With the speed of innovation in the technology of data collection and communication (both in advances in sensor technologies as well as data transmission and security capabilities),  the biggest limitation to the benefit of these new technologies may well be a lack of expertise to fully utilize the surge of available data. It is interesting to note that a recent study reported that only 10% of the data being collected off the shop floor is usable. As more sensors are installed in more places, it is expected that this figure will quickly climb to at least 30%. The challenge will then be capitalizing on the benefits of this data. Too few engineers are adept at understanding what data to use, what not to use, and how to bring it all together collectively to determine the real root-causes of problems using such tools as Minitab to assist with statistical analysis, let alone how to model comprehensive automated system adjustment algorithms that make use of the increasing amount of available data.

Going forward, it will be important that engineering teams expand their capabilities to include the design of modeling systems that will take the wealth data from throughout the factory to a computerized conclusion. Engineers of all types will need to have a much more thorough understanding of data analysis and usage that extends further into algorithm development, simulation modeling, and automated systems control.