Martin Høgsbjerg, Head of Supply Chain Technology in Denmark, at Danfoss Drives

Martin Høgsbjerg, Head of Supply Chain Technology in Denmark, at Danfoss Drives

There are many contributing facets behind the increasing use of robotics in the factory, but three of them stand out, in particular:

- Healthier working environment for staff

- Improved quality in the manufacturing process

- Reduced manufacturing costs

At Danfoss, we invest extreme efforts in achieving these benefits using robotics. And to do that we have overcome some big challenges.

There is no easy way to establish and run a world-class manufacturing facility. It demands constant attention to new development and active participation in international robotics forums and development projects, to ensure Danfoss is always operating at the cutting edge of automation technology. This effort is also required to ensure that we implement automation at a responsibly safe level of maturity. In collaboration with international experts, we constantly keep up with the reliability limits of automation and apply them in the best way possible to our production of AC drives.

"DfAA is a clear example of the close collaboration required between R&D and the factory"

The best thing about robots is also the worst

Robots perform exactly the same action every time. This is where their value lies, in reliability, precision and standardization. However, manufacturing processes that involve a high degree of variance and non-standard steps, performing exactly the same action is a liability for the robot, meaning it can have difficulty performing its tasks. At Danfoss how do we help the robot overcome the challenge of handling and mounting a huge variety of different-shaped components? We apply the science known as Design for Automated Assembly (DfAA). What concrete steps do we take using DfAA?

Firstly, we design our manufacturing equipment to assist the robot: we design in mechanical fixtures and sensors to help the robot make decisions. For example, a vision sensor can help the robot pick up components correctly. Secondly, we design our product components and packagingin all kinds of detail, to preempt or compensate for possible variations in tolerance: 

We use common gripping elements on different components, wherever possible. We put reference guides in place, that gently tap and rotate components into the correct position. All these efforts are examples of self-alignment design. This approach to design is an iterative process and requires ongoing negotiation between the factory and the R&D department. Naturally, we don’t want to compromise on design where it might affect product performance. It’s a sophisticated balance requiring advanced tools to simulate the consequences of each individual decision.

Collaboration between manufacturing and development

DfAA is a clear example of the close collaboration required between R&D and the factory. This collaboration is illustrated in the hierarchy of criteria of product characteristics, which forms the basis for manufacturing priorities.

The hierarchy of criteria includes characteristics which we want our products to exhibit, ranked in level of importance. Characteristics which find their way into the hierarchy include quality, size and footprint, commissioning complexity, dependency on packing format, ergonomics, scalability, robustness and more. Then based on the hierarchy, we define and prioritize manufacturing principles, such as

- Clean and simple production flow

- Dedicated areas for kitting, assembly, testing and more

- Separate aisles for trucks and stackers

- Continuous flow with a minimum of buffers

Flexible and robust manufacturing

Traditionally a manufacturing process is built up on a line basis. However, to build robustness, the smart factory is built up of multiple standalone cells, which perform defined tasks. These cells are then connected intelligently to ensure a logical workflow where the drive passes through several cells during its assembly. At every step there is more than one alternative route, so that in the event of a delay at one cell, the manufacturing path progresses to an alternative cell which can perform the same manufacturing step.

The flexibility and many degrees of freedom that the flexible cell concept implies, means the process can quickly become very complex from a production planning point of view. Therefore, each manufacturing facility must choose its own compromise between flexibility and simplicity and cost-effectiveness: the more flexibility, the more complexity. Simplicity can still be a virtue and there is every reason to implement just enough redundancy and process robustness to ensure a reliable manufacturing capacity, but no more.

To optimize this balance between flexibility, simplicity and cost, the factory runs simulations to estimate losses and gains prior to establishing the equipment in real life. The simulations provide the information base required to make the decision on the level of complexity required to achieve a robust process.

In practice, the new factory is built up one cell at a time, scaling up to meet higher demand over time.  Artificial intelligence controls the network of machines to achieve optimal flexibility.

Bringing even more robots into the factory

In the past we have only automated our PCB production and certain sub-assemblies in manufacturing using robots. Now, we are going through an automation transformation of the whole global supply chain. We have started the journey with the highest-volume items, and with assembly and test areas.

Many vital contributors

There are many contributors to successful smart manufacturing, and we have discussed several of them here. As you would expect, connectivity, communication and artificial intelligence are important elements. The less visible elements are compromise and alignment between product design and manufacturing processes, a matrix approach to process robustness to ensure a reliable manufacturing capacity, and optimal packaging of both components and finished products. All these contributions play a special role in smart manufacturing – each of them vital –to create products of world-class quality and reliability.

In the future, robotics is a vital element of our plans. We aim to make step-changes in safety, quality, cost, and delivery for the benefit of our customers and staff. These goals fit perfectly into our ambition of maintaining our world-class position inflexible supply chain operations.

Robotics

Supply Chain

Sensor