Applications and inventions, quoting custom machinery

This article is intended to explain a little bit about the applications process machine builders use when quoting custom machinery.

By Frank Lamb, Automation Primer February 11, 2015

Just because a machine is customized to produce a specific product doesn’t mean it is unique or necessarily qualifies as an invention. There are a lot of standard subassemblies that can be combined in creative ways to make things. Some examples include robots, conveyors, dial tables, pick-and-places and indexers. On a smaller scale sensors, servos, tooling and mechanisms are mixed in innovative ways to accomplish different tasks, all connected by a control system. Even though the resulting machine may look unique because of the way it was built, machine builders would consider this fairly standard stuff. Applications engineers have seen a lot of types of systems and come up with lots of different, but normal ways of doing things.

Sometimes an application can have a challenging aspect that requires a designer to think outside of the box. A common situation can arise where a product is made manually on workstations by human operators. The human hand, eyes and brain work together to do things that machines find very challenging. Robots are stronger and faster, but are generally limited to repeating the exact same motions over and over. Sensors can be used to change these motions, but the decision making capability of the human brain and the coordination of the eyes and hands combine to do some amazing things.

When these applications arise, innovative ways have to be used to solve the problem. Even if someone has solved the application before, the designer may not have seen the solution and must come up with their own idea. This is where patentable methods and inventions come into play. While the human eye and hand can accomplish amazingly dexterous tasks, there are things that mechanisms and machinery can do much better. For instance, the hand can only rotate around 300 degrees or so at the wrist, whereas a mechanism with a gripper can rotate continuously for many revolutions. A human operator can only hold two hands worth of components at one time, whereas the number of items that can be positioned together by a machine is limitless. Identifying the method that will be used for processing, production or assembly is the crucial task of an applications engineer or inventor.

This can create a dilemma for a customer trying to find a way to build a non-standard production machine. It costs quite a bit of time and energy to put together a quote. Larger machine builders have a department dedicated to applications and quoting with multiple experienced people that can bounce ideas off of each other. They can also draw on resources from their mechanical and controls departments if necessary. The problem is that these larger machine builders have higher overhead and don’t usually build smaller machines. It is simply not economical to do so unless there is potential for multiple machines. At the same time, smaller machine builders or machine shops who might be interested in building a lower cost, smaller machine don’t have the application resources or experience to spend time coming up with innovative ways to do things.

One method of proving out an idea is to build a mechanical “Proof of Principal” (PoP) machine. If something is being built by hand at a workstation, sometimes the station can be modified using various mechanisms that can be manipulated by hand by the operator. Once the concept has been proven out mechanically, mechanisms and actuators can be used to sequence through an operation at higher speeds. Sometimes these PoP stations can be later cannibalized to provide parts for the final production machinery, saving money and providing smaller machine builders or machine shops with a baseline to design around.

How do you handle preliminary design and concepting for machinery? How about risk management and PoPs? I would love to hear from both machine builders and manufacturers on this subject.