Modern CAD tools and simulation software allow engineers to analyse much of a machine’s behaviour before anything is physically built. At the same time, machines are becoming increasingly complex. Alongside mechanics, automation, sensors, control systems, software and data now play a much larger role.
This raises an important question in modern development: do we still need physical prototypes? According to Revismo’s Head of Machinery, Mirko Arras, the answer is yes — but the role of the prototype has changed considerably.
Mirko has worked as a lead engineer at Revismo for 13 years and has been involved in a wide range of machine and product development projects. Revismo has contributed to more than 850 development projects in total. Based on this experience, the physical prototype remains an important part of development, but much more engineering can now be completed before the first machine is built.
„Today, the first physical machine should already be quite close to a working solution. We simply have many more tools available to verify the design and operating principle before the machine is built,” Mirko explains.
That does not mean the first machine must work perfectly as soon as it is switched on. Almost every new machine requires some adjustment of speeds, sensors, timings, forces or control logic. The important difference is whether engineers are tuning a fundamentally sound solution or only starting to solve major mechanical and design problems once the physical prototype has already been built.
In our development projects, we have never ended up with an unusable prototype. Every prototype has ultimately become a working, productive machine,” Mirko adds.
A Good Prototype Starts with a Good Specification
According to Mirko, this leads to one of the most important questions in product development: how well has the project been thought through before engineering begins, and how clearly has the actual problem been described to the engineering team?
A clear and accurate specification is the foundation of a successful development project. If, for example, a component is specified as being 100 mm wide, but in reality, its dimensions may vary between 90 and 115 mm, the engineering challenge is already significantly different. The same applies to production volume, cycle time, material properties, operating environment, cleaning, maintenance and safety.
For automated equipment, there are also questions related to machine control and data. Does the machine need to communicate with an existing production line? What data needs to be collected? Does the equipment require different recipes or operating modes? How easy should it be to modify the machine later or add new products?
“The more important questions we can resolve at the beginning of the development process, the fewer surprises there will be later. An engineer can solve the problem that has been described to them, so the quality of the initial information should never be underestimated,” Mirko says.
Modern Machines Need to Be Engineered as Complete Systems
One of the biggest changes in machine engineering is that mechanics, automation and software can no longer be considered separately – and increasingly, they should not be.
In some cases, a complex mechanical solution can be replaced with a simpler structure, a sensor and smarter control logic. In other cases, a mechanical solution may be more robust and practical than adding more sophisticated automation. All these choices must also be assessed from a safety perspective.
According to Mirko, this is where an experienced engineering team creates real value: by assessing these trade-offs at system level. The question is not simply whether an individual part can be designed or whether a suitable sensor can be selected. The key is how all elements work together so that the machine is reliable, manufacturable, maintainable, safe and fit for its intended purpose.
Manufacturing and assembly quality are just as important as the design itself. Building a complex machine is not simply a matter of assembling components according to drawings. The result depends on assembly accuracy, alignment, commissioning and the integration of mechanical, electrical and automation components.
This is why building a complex prototype requires specialists who understand not only the individual components, but also the overall operating principle of the machine.
Not Everything Can Be Predicted on a Computer Screen
More capable software does not eliminate the need for a physical machine. One reason is very simple: the real world does not always behave exactly as the input data suggests.
This is particularly important in machines that handle natural or otherwise variable materials. A specification may define a particular size, weight or other property, but in real production the material is rarely completely consistent.
“The material being processed often has variable properties. Dimensions may vary, the material may be drier or wetter, or it may simply behave slightly differently from what was assumed when the original specification was prepared. We can predict a great deal digitally, but ultimately, the machine still has to perform reliably with real material.”
Engineers can account for many of these variables during design, but it is not always practical or even possible to simulate every possible combination digitally.
This is exactly where a physical prototype remains valuable: it allows the engineering team to verify how the solution performs in the real process.
Tuning Is Not the Same as Redesigning
Some degree of tuning is normal with a new machine and, with modern automated equipment, is often unavoidable.
Speeds, accelerations, sensor sensitivity, timings and other parameters frequently require final adjustment under real operating conditions. The important distinction is between tuning and redesign.
In a well-executed development project, the basic structure, operating principle and selection of key components should already be correct. Once the physical machine is built, the focus should be on finding the optimal operating point.
The data collected by the machine can make this optimization increasingly precise.
“Most machines can eventually be made to work. But if enough engineering has been done at the beginning of the project, the prototype stage should mainly be about tuning and optimization.” says Mirko.
Physical and Digital Development Are Moving Closer Together
Mirko does not see the future of machine engineering as the disappearance of physical prototypes. Instead, digital models, engineering calculations, simulation, automation, machine data and physical testing are becoming increasingly integrated.
“Advances in software have not made physics disappear. A real machine still has real components, tolerances, materials and a real production environment. Good product development means thinking through as much as possible digitally and then using the physical machine to verify everything that the real world adds.” explains Mirko.
The question, then, is no longer simply whether a physical prototype is necessary. The more relevant question is: what should we expect from a prototype today?