Batch size one: Small batch sizes becoming cost effective—even for small shops
Improvements in manufacturing management software, robotics, additive manufacturing and thermal controls are making small batch sizes more cost effective—even for smaller shops. Manufacturing plants are able to reduce inventory, improve throughput and reduce demands on human operators.
“We’ve had more customers coming in looking to do smart parts, just batch quantities,” said Andy Wiese, integration supervisor at Acieta. “For certain customers, parametric programming and offline programming is really their only option if they want to do these small-batch jobs.”
Accurate motion and temperature control often are critical in five-axis machining of 3D parts in small batches, said Gisbert Ledvon, director of business development at Heidenhain. For some high-precision, injection-molded parts, such as those used to mass produce airbag covers or dashboard panel components, accuracy must be within microns, he added.
But the temperature range on a factory floor in the Midwest can vary from 68 degrees F in the morning to 80 degrees F in the afternoon, Ledvon said. Even a one-degree variation affects precision.
“Most machine shops in the US have to deal with heat and temperature deviations throughout the day. If you have a temperature variation of only one degree, one meter of steel can have an expansion of about 11 microns,” he said. “There’s nothing you can do about it other than cooling the machine or cooling the entire environment.”
Without Heidenhain’s KinematiacsOpt, an operator would have to come in to manually verify the machine was positioned properly, Ledvon said.
The technology really proves its worth on parts like big mold bases that are already worth thousands of dollars where accuracy is critical.
“If you want to hold accuracy to within a few microns, you really need it,” Ledvon said. “Before that part goes on the five-axis machine, the customer may have spent 30 to 50 labor-hours with a lot of work in drilling, pre-machining, heat treating and grinding. The customer has only one shot to cut it right; otherwise, they’ve wasted thousands of dollars and man hours.”
Many Heidenhain customers in Michigan and in Windsor, Ontario, manufacture both large and small injection molds in small batch sizes, he said. Before using Heidenhain’s five-axis motion system with closed loop thermal control, the process was more labor intensive and expensive.
One application in automotive is to make molds that will be used to make airbag covers, Ledvon said. The molds will be produced in small batch sizes to be used in auto plants all around the world. All of the molds need to be exactly the same.
“You produce one mold and then you make a million covers,” he said. “They need to make sure that the molds are true to size. When they’re assembling the cars, there might be parts from Mexico or from Germany to put in that car. You need to make sure these components look exactly the same.”
Production optimized at the factory level
With automated dynamic scheduling systems, such as Fastems LLC’s MMS7, companies can optimize production at the factory level as opposed to each work cell, President David Suica said.
One global tooling company saw throughput improve three times, he added.
In the past, employees in a work cell would process whichever order arrived first, Suica said. Dynamic scheduling systems, however, set the order of operations based on the urgency of orders and when parts need to be inspected.
“The second order may be the higher priority,” he said. “The MMS7 running on the operator’s PC assigns priorities and tells the operator which order has the higher priority, which changes all the time.”
At another company, scheduling had been based only on oven utilization for heat treatment, Suica said. Since smaller parts require an hour in the oven and larger parts take three hours, the company usually scheduled the small parts first. But the scheduling didn’t consider when the parts were needed in production.
“They optimized the oven but not the process. Running one cycle of the oven was cheaper for the oven but now you have all this inventory sitting around,” he said. “They had local optimization but not optimization for the best ROI throughout. The system dynamically reschedules your shop. You’re automating the control and flow of the shop.”
For example, the system first instructs a human worker to load raw material into the machine; the next step is to make sure certain tools are at hand; then the worker determines that the right program is uploaded; and finally, the worker is cleared to begin the next step, Suica said.
In the meantime, the MMS also is scheduling the previous order to make sure it’s off the grinder by the time this order needs to be on the grinder, he said.
“Lesser skilled people can run a line with dynamic demand and response,” he added.
Rush jobs are no problem
Speaking of scheduling, responding to rush orders—no matter the batch size—is much easier with a dynamic scheduling system in place, Suica said.
Instead of a line manager running around frantically with a clipboard to each work cell, the system automatically adds the hot job and reschedules the routine orders.
“Now if you get a rush order for 50 parts, you could build those 50 for that customer today instead of having to have 50 parts on the shelf ready to go for that customer,” he said. “You’re building to demand instead of building to plan.”
Heidenhain also offers a batch process manager that easily enables rush jobs, Ledvon said.
“If you’re in the middle of production and the customer calls and says, ‘I don’t need this part yet. I need this part instead,” with our batch process manager, our controller will allow you to change on the fly,” he said. “The controller will automatically look in the program and make sure you have all the cutting tools available for the new job. Our controller will say to the process manager, ‘Push this button and you can cut this part immediately.’ It gives you a lot of flexibility to help your customer on demand.”
Plays well with others
Interoperability is another area where improvement has been marked.
“When people talk about which controller to use, it’s not just how fast the controller is but how easy it is to connect to the rest of the system,” said Keith Kersten, industry marketing group manager at Omron Automation Americas. “What used to be found only in high-end controllers, you now can find in more entry level controllers. We offer connectivity to outside systems and databases in a wider range of controllers.”
The best integration involves more than one machine talking to another but the ability to communicate among different vendors and technologies and from the factory level to the enterprise level, he said.
Teaching robots and people is easier
In the past, adding a new part required a robot operator to teach new points on each robot, Acieta’s Wiese said.
“You had to physically copy the programs and step through each process for that part. Before, you could spend half a day teaching four robots a new part,” he said. “Now, the operators can simply enter the length, width and height of the part and the system runs. With this system, we write the code and set up the system ahead of time. When the customer gets it, the code is already there and the operator just has to enter the dimensions.”
Overall, systems are becoming easier to use and integrate so that control engineers don’t have to be control experts and IT users don’t have to be engineering controllers, Kersten said. “You can transfer data between IT and OT without jeopardizing real-time flow on the controller level,” he said.
Improving technology has made adding robots to a manufacturing line easier and more cost effective for smaller plants, Kersten said.
For Acieta, these advances mean the company can offer its clients a standard human-machine interface that cuts lead times on standard products to one week or less and custom solutions that average 20 weeks, Ryan Morgan, senior mechanical engineer, said.
“The customer buys this standard system,” Wiese said. “The system leaves our facility ready to weld. Once they get the system, all they have to do is put in the fixture, load in offline weld programs, and it’s ready to weld. It has improved our ability to serve new robotic users with a lower price and faster delivery options. We are able to introduce robotics to more manufacturers who in time may choose to move to a larger custom system once they see the benefits first-hand.”
For example, customers in Arkansas and Michigan that make conveyor beds and electric boxes of varying sizes, respectively, are able to easily customize factory runs, he said.
“They could have an infinite amount of sizes of parts” Wiese said. “If they had a system that couldn’t be run batch style, they would be teaching parts all the time. This style of program allows them to run any size of parts they need instantly.”
Better simulations help
Along with improvements in robots and cobots, better simulation programs have helped manufacturers develop motion planning for advanced robotics systems.
That’s according to Tom Hoffman, portfolio development executive at Siemens Digital Industries Software.
“These software solutions provide the ability to quickly program alternatives that then can be loaded into the system without negatively affecting currently running production,” he said.