High-tech advanced manufacturing: interview with Flanders Make CEO Dirk Torfs

High-tech advanced manufacturing: interview with Flanders Make CEO Dirk Torfs

A major objective of Flanders Make, the strategic R&D centre to develop new manufacturing technologies, is to advance Industry 4.0 objective through automation, robotics, algorithms etc. Its CEO since 2014 is Dr Dirk Torfs, a distinguished civil, mechanical and electrical engineer who also holds a doctorate in Applied Sciences. He is leading a high-performance research team who is tasked to develop groundbreaking manufacturing processes aimed at achieving competitive edge through innovation that is both incremental and disruptive. The aim is to create factories of the future through open innovation. Flanders Make is committed to developing new products and production processes. It operates from three sites in Lommel, Leuven and Kortrijk and from 12 research facilities at Flemish universities. In this interview Dr Torfs speaks about manufacturing trends that will add to competitiveness.

What is the current state of thinking among engineers and management on industrial automation that aims to minimise manual intervention in operating machine equipment or production process?

From a management perspective, industrial automation is the enabler and driver to achieve predictability, reliability and zero defects in manufacturing. Industrial automation is the only way to boost productivity and fulfil the need for customised, smart, interconnected products, delivered on the spot (very short lead times) at a cost similar to mass production.

This is possible by using machine equipment that takes autonomous decisions during the operation. Industrial automation solutions capture the operator’s knowledge or provide high value support to operators via Artificial Intelligence algorithms and big data analytics.

The availability of sensors and high performance processors lead to a detailed analysis of the production process and the machine status. This ensures (1) that the production process can be quickly tuned for the required task and (2) that the machine can be adapted depending on factory-specific working conditions. Behaviour of machines becomes very predictable such that defects can be predicted allowing minimal downtime. In future, digital twins of physical products provide answers to questions such as “what is the remaining lifetime” or “when is the next failure occurring and on which component”.

Digital technologies support the supervisor in optimising the production process leading to zero defect production and shorter lead times. Production engineers or operators might have a different perception upon the value of industrial automation. Acceptance of technology is often overlooked. As a consequence, people can feel threatened and fear for their job. This is because not enough attention has been paid on getting the technology accepted before the implementation. It makes much more sense to implement an automation process in different steps:

  1. limited automation based on full control of the operator. He offers valuable assistance for optimising the machine or the production process.
  2. basic tasks are fully taken over, but special tasks are still under full control of the operator. More and more information is made available as decision-supporting content.
  3. the machines are put in full automaton mode.

As a result, operators gradually realise that they are unable to perform that specific task at the same level as the machine. Moreover, the operator regains time for more valuable tasks making his job more relevant.

The automotive world is an engineering intensive environment. New technologies are considered as the playground by many employees, giving them a very satisfying job and allowing them to be creative in developing new functionalities corresponding to the needs of the end-user.

human - robot collaboration

Through the application of computer, electronic equipment, control theory, and related process technologies, industrial automation is being introduced widely. How challenging is data acquisition and data mining in running the manufacturing process functions to obtain optimisation, detection, control, and regulation of the whole industrial production process?

Data is the new gold, but hardly anyone knows how to capture value from all that data. We identified the main issues.

  1. Data acquisition on the production floor, although considered to be easy, requires a coordination of IT (information technology) and OT (operation technology). The main challenge is to have a common understanding about the way data is measured, structured and stored such that data mining becomes possible. Data that is now generated by different departments may not necessarily be stored using the same storage mechanisms.
     
  2. For companies to optimally benefit out of the digital transformation, an organisation and leadership change is often needed. OT should take over the lead from IT/ICT. Today IT is leading in many companies safeguarding standards, restricting access and overruling OT on decisions related to tool selection, implementation and data storage and usage.
     
  3. To optimally use data for decision-making, it is necessary that OEMs have access to data of the final product/machine. This access enables them to give valuable advice for optimising the use and set-points of the machines on the production floor. However, security concerns lead to limited external access to company data networks. Consequently, data mining can only be done without the background/inside of the production application. The valuable advice is, therefore, less effective.
     
  4. Production environments need reliable, predictable and explainable decisions and results. Therefore, there has to be trust in new technology that supports decision-making. One Uber accident leads to a significant delay in the social acceptance of self-driving cars, although objectively the impact of human errors is much higher.
     
  5. A final issue is the computing power needed by artificial intelligence algorithms. Companies need to develop a strategy for AI calculations. Cloud computing is powerful but slow, since data transmissions rates are not guaranteed. Edge computing is less performant but more stable due to the lower bandwidth needed. Device computing is even less performant but very sure as latency is not an issue. 5G will significantly improve the access to computing power.

An example from the automotive sector of technology used to improve or develop new products:

Traditional suspension systems are passive, spring-based and reduce the impact of the road on the comfort of the driver and its passengers. For race cars, it was soon acknowledged that active suspension systems increase the stability and safety of the car. As a result, semi-active or reactive systems came into the market for passenger cars as well. Sensors were introduced and on-board processing was made possible. Today, fully active suspension systems are entering the market. Sensors measure and evaluate the road ahead (detecting pot-holes, for example) and adapt the behaviour of the suspension system. The driver doesn’t feel the pot-hole and comfort becomes independent of the road quality.

In the tyre and automotive industry how far has the world gone in the successful introduction of Industry 4.0 practices? What are the problems that are being faced by existing industry as it struggles with intensive capital investments that are needed to replace ‘legacy’ systems? Are such costs justifiable?

PricewaterhouseCoopers has shown in a recent study that the digital transformation has a lot of potential, but is still rather limited in its execution. These results confirm earlier studies by Flanders Make and PWC. Asia Pacific (APAC) is clearly taking the lead with 19% of digital champions, while Europe, Middle- East & Africa (EMEA) and the Americas have a significant delay leading with respectively 5% and 11% of digital champions. As the same trend is present for the digital innovators, a catching up of the lagging regions is not to be expected.

The positive news is that in the automotive sector 20% of the companies can be classified as digital champions and 34% as digital innovators. Hence, about 54% of them are soon reaching digital maturity, allowing them to fully capture the value of Industry 4.0.

It is not surprising that digitalisation is more mature in the automotive sector. Many supporting technologies (context sensing, camera’s, sensors on all subcomponents, data and communication backbone etc) are already being implemented to optimise the customer experience: adaptive cruise control, lane and pilot assist, city safety, active suspensions systems, eco driving, etc. Cars are becoming smart, interconnected devices.

In addition, new services connect the car manufacturer with the end-user. Firstly, via charging stops for electric cars. Secondly, via the Internet through advanced functionalities such as automated emergency calls or car assist. This results in massive data records of how cars are used, allowing new digital services adapted to the individual customer’s profile and expectations.

A nice example is HERE WeGo: a web mapping and navigation service, operated by HERE Technologies. HERE Technologies is a company that provides mapping and location data and related services. It is majority-owned by a consortium of German automotive companies (BMW, Audi, BMW and Daimler). They collect data of the use of the car which is very valuable for future car development.

Finally, it is true that investments in digitalisation are high and could be a barrier. However, a sustainable and long term investment plan based upon a roadmap for roll-out of the digital strategy will lead to results and revenues. Digital champions are already referring to a revenue percentage of 24% today and 36% in five years. Traditional products and services are expected to decrease to less than 30% in five years from now. Not investing is not an option.

Collaboration with other companies can be helpful to spread investments and create a strong backbone for capturing value. A good example is autonomous driving. Premium brand manufacturers such as BMW and Mercedes are joining forces to create self-driving car solutions.

How should CEOs prepare their firms to switch to intelligent manufacturing that calls for intense digitalisation, networking, and integration?

CEOs should prepare their organisation and their people for the change that digital transformation requires. Having or installing an appropriate/adapted culture in the company is key to the success of any new strategy. This often overlooked. The ecosystem for innovation, implementing Industry 4.0, consists of a number of strongly interlinked building blocks which are unlocking the company value. They require a few points of attention.

  1. Managing product and production is the first challenge. Companies tend to focus on either product development or production. This leads to inefficiencies. Both should be considered in parallel.
     
  2. Secondly, companies tend to invest more in R&D with long term perspectives for products. Production should follow the same trend but is systematically considered as a short-term investment with an ROI of preferably a few months. This is why many companies have only one lead factory instead of many. A good production strategy for reaching and staying a lead plant is a strong driver for staying competitive in the market.
     
  3. Third, a local and efficient production, close to the customer, is becoming increasingly important since digitalisation allows for a significant reduction in lead times. A good digital strategy starts with a sound analysis of the local market potential and the definition of an appropriate (often new) business model.
     
  4. It is essential to have a culture of agility (adapting to the new world, needs and technologies) and collaboration between departments as well as with external parties completing the value chain and knowledge actors. The in-flow of new ideas accelerates knowledge development and leads to win-win-win situations.
     
  5. Finally, people are very important in the unlocking of the value of digitalisation since the use of technology (for products, within the production environment and in the market) depends on people. But the skills and culture of both employees and users can hinder the roll-out of an effective Industry 4.0 strategy.
     
  6. Technology and research are, in addition to people, strong enablers for unlocking the potential of Industry 4.0 during the digital transformation journey. Setting up broad and strong collaborative innovation partnerships and if possible ecosystems, of which the size will depend upon the specific IP requirements and needs, is needed in a VUCA (volatile, uncertain, complex, ambiguous) world.

To conclude, it is important to highlight that reaching an high level of digital maturity requires to excel and accelerate at all domains equally. Good products with a non-collaborative culture or a not appropriate business model will lead to inefficiencies. The investments will not yield.

Dirk Torfs, CEO Flanders Make

"The automotive world is an engineering intensive environment. New technologies are considered as the playground by many employees, giving them a very satisfying job and allowing them to be creative in developing new functionalities corresponding to the needs of the end-user."

Dirk Torfs

CEO - Flanders Make

What are the security issues involved as the deployment of Internet-of-Things and cyber dependence creates security issues that could impact the manufacturing process?

Connectivity can create issues when no measures are taken. Current IT systems are well-protected against security issues and standard Internet is generally not used for industrial purposes. The main issues arise when foreign devices enter the manufacturing floor. This is often prevented.

Consequently, remote functionalities are reduced and feedback from digital servicing by the OEMs becomes impossible. The data they capture, however, could be valuable for data analysis, optimising the product’s/production’s performance through condition monitoring.

Companies need to make trade-off between potentially opening the network to external threats and optimal performance thanks to remote functionalities and decision-supporting data.

How would one address the issue of labour in the context of likely reduction in employment with trade unions opposing to what they describe the ‘dehumanising’ of factory floors, with robots taking over most functions of the workers?

I strongly believe - and the first reports on the impact of industrial automation confirm this - that digitalisation will not significantly impact labour. I am sure that it will displace some jobs and that shifts between jobs will occur. But more importantly, digitalisation will lead to the re-skilling of existing talent.

Since the 1980s, robots have increased productivity and induced changes in the workplace (making them more safe), enhancing welfare and creating new ‘high-skill’ jobs. Technology makes work more attractive since repetitive jobs can be replaced by empowering jobs. Technology has and will improve work by complementing human capabilities.

For example, digital work instructions and augmented reality projection systems support operators

  1. by providing tailored instructions based on experience level
  2. by (remotely) supervising the execution of the task to ensure zero defect production
  3. by measuring the physical/cognitive impact of a task and consequently optimising the distribution of the tasks between robots/cobots and operators

In this way, technology empowers people, creates high quality production, enhances their capabilities and increases the satisfaction level.

Contact us

Do you want to know more about the Make Lab and the possibilities for your company? Do not hesitate to contact us.

Dirk Torfs, CEO

Dirk Torfs is CEO of Flanders Make since 2014. Dirk is a Civil, Mechanical and Electrical Engineer as well as a Doctor in Applied Sciences (KU Leuven). He has over 20 years of experience in management positions in the Flemish industry and is Professor of Quantitative Decision-Making for the Executive MBA programme of the Flanders Business School.