Co-Innovating the Future: Yokogawa Accelerates Progress towards AI Adoption

October 7th 2020

Artificial intelligence. We hear about it every day and sometimes it still feels like a term that belongs to the future. But is that the case? Because even a simple smart device can already contain a - perhaps small - variant of AI. Simply put, AI is the automation of mind power. How can this automated power of the mind benefit us? At Yokogawa, Marco Bischoff and Martijn Kramer gather around the table to share their vision and experience around this subject. The sequence we follow is ‘people, process, technology’. If you supply pressure transmitters, it might be less important; but if you supply solutions where the customer does not only want data but useable insights - and algorithms generate such insights in an intelligent way – that requires an extra effort from the organization and their employees. 

 Yokogawa is a well-known organization in the OT world – famous for its sensors, DCS and SCADA systems. The company also started to develop towards the IT/OT side - MES, Operations Management, Security, etc. - and Business Management. In the past years, a brand new IoT program for the industry is developed. The overarching vision on all these subareas is to develop with the customer together towards one direction: Autonomous Plants. 

Having experience in various business units, Marco and Martijn are both involved in the development of the new IIoT program, whereas AI is an important part of it. Marco is Product Manager Process Control Instrumentation and Martijn works as Senior Consultant Advanced Solutions & Innovations. The business unit where Marco works, Network Solutions, had an open structure right from the start. However, the open structure has now affected the company as a whole. "We used to be more concerned with putting our products on the market. Nowadays, we’re more listening to our customers. We are in a movement from products to solutions, together with our customers. After all, every customer is different as well as their situation. Every client needs something different. Marco laughs enthusiastically and can talk for hours about Yokogawa, the development and the structure of the services. "Our slogan 'Co-Innovating Tomorrow' is very recognizable – we collaborate with the customer to develop new solutions together."

"Using the information intelligently ensures that solutions can communicate with each other", Martijn adds. He sees the effects of that change within the organization. "There used to be more pillars in the company, each dealing with their own products and solutions. Nowadays, you notice that solutions are being developed throughout the company, with all disciplines together. As a result, we as employees are more broadly oriented and we learn more from each other". Moreover, customers are also changing. "We frequently see different disciplines on the customer side. One question can simply no longer be pigeonholed, it’s an organic process”.

The added value of AI
In many aspects, the added value of AI can be found. The use cases now mainly focus on the following applications: the detection of abnormalities for predictive maintenance. For example: data from IoT sensors such as the Sushi Sensor, combined with AI-based algorithms, can predict in a timely manner when assets require maintenance. This allows you to take timely measures and schedule maintenance before the minute deviations you detect, leading to a more measurable deviation in your process. This way, you can greatly increase the availability of your production process.

Error analysis, or root cause analysis: changes in the process or in the condition of machines, for example in the field of quality or energy consumption, can be traced back to the cause. By tackling those causes, you can significantly improve your process.

Quality control: without time-consuming lab checks and/or destructive tests, you can predict the quality of your end product, make quality classifications and anticipate deviations in quality - and thus be ahead of the game, with no extra cost. 

Due to unplanned delays and downtime, a lot of capacity - and thus turnover - is lost annually, not to forget to mention the energy loss. Sustainability is also becoming an increasingly important driver. Gradually, these developments lead to improvements. Eventually, this will translate into the realization of autonomous installations.
 

From Programming to Decision-Making
Marco: "A classic automation system is programmed step by step. With the rise of Industry 4.0, decisions are based on data. This is a major change in the automation concept, with major customer demands. 'Allowing my process to depend on something in the cloud’? That feeling that it is elusive and that you are relinquishing control is, of course, very exciting. You might raise the same questions as before. Wireless? Is it reliable? My data in the cloud? Is it safe? Will I not lose control? Thus, it is important to do everything together with the customer".

In 2015, the New Technology Center department in Japan started developing AI with the aim of being able to use this to eventually achieve autonomous plants. From industrial automation to industrial autonomy. "Yokogawa has developed algorithms with which we create industrial value for the customer. Of all the data that sensors transmit, only a small percentage is valuable for a specific application: in the event of deviations. You need tools to extract the right data and analyze it in order to make decisions. Many people still have to get used to new ways of working. You can easily start monitoring a health score. To give an example, freezers are commonly used in pharmacy. The course and stability of the temperature signal says something about the condition of the compressor". Marco laughs: "In fact, just like how my father-in-law can tell the status of the car combustion from the sound and vibrations of its engine. The knowledge and experience the experienced engineer uses to say something about a running machine can also be found at AI".  This control often remains a tricky issue. Having everything 'on premise' often feels safe, but is that really necessary? But a shift can happen very quickly, especially under the current circumstances. At the beginning of this year we couldn't really imagine how we all work together now. 

One-stop-shop
 "Open systems are beneficial to many aspects and this is what Yokogawa strives for. Up until now, we launched the IIoT program JOIN where we work together with both customers and partners on the solutions. We partnered up with Perfact Group, a powerful partner in the field of maintenance that enables us to offer a one-stop-shop in the field of Industry 4.0, across the entire value chain of consultancy, data collection, connectivity, cloud, forecasting - using AI and dashboards".

Martijn opens a little kit with the new IIoT Sensor. "An important contribution to such intelligent solutions is the Sushi Sensor. A small, robust, high quality sensor. We have taken a clear direction in the field of IoT sensor technology. The Atex-certified Sushi Sensor can perform various measurements, depending on the type. With a strong magnet or a stud, the sensor is attached to a device and wirelessly measures vibrations and temperature, which are sent via a LoRa transmitter via the cloud to a dashboard. Recently, two other varieties of the Sushi Sensor are launched: The Wireless Temperature Sensor and the Wireless Pressure sensor are launched. You can use it for a great range of applications: think of a vibration measurement on an engine or a pump or monitoring the degree of contamination of a heat exchanger". 

Sushi sensor: craftsmanship
The Sushi-sensor is specifically developed for applications that are located in harsh conditions in the industry and collect a huge amount of data. "We offer customers a starter kit with three sensors, a LoRa gateway, access to the cloud. and various algorithms that turn all that data into usable information. In the meantime, many use cases are in development so we can support our customers in a concrete way. The application enables us to identify errors timely and prevent shutdowns. Bottom line: preventing downtime, producing more efficiently, and operating more sustainably. The knowledge to apply the sensors, using a LoRa network, is available. This way we can support customers to shift their maintenance work from reactive to prescriptive. This starter kit contains everything you need to set up your own IoT project effortlessly and at a relatively low cost". Doesn’t this seem like the perfect opportunity?

Martijn shows various application examples from real-life practices. "By co-innovating with our customers, we offer an appropriate solution for these challenges. AI plays an increasingly important role in this!  We actively invite our customers to join us to gain experience in this field".
All predictive models about the speed and impact of the digital transformation and AI are likely to be up and running by now. McKinsey & Co, for example, previously researched into that impact. Can we cautiously estimate that it will be bigger than previously estimated?
The need for further automation with a high degree of autonomy is greater than ever. AI is here to stay.

 

Sushi-starterkit
Three versions of the JOIN starter kit are launched recently. Depending on your application, you can order an asset-specific kit. You will have a complete solution for heat exchangers, pumps or hand valves. The measurements are sent to a dashboard via a LoRaWAN connection. Let’s see a few examples of applications that were equipped with a Sushi Sensor. 

Sushi Sensors were mounted on various pumps and collected hourly data of the surface temperature, vibration velocity and acceleration for six months. As the acceleration value increased and the curve changed, the pump was disassembled and the bearing was found to be damaged. An unexpected shutdown of the installations was prevented. On another installation, the Sushi Sensors were mounted on pumps to measure vibration, acceleration, speed and surface temperature. In the past, a field operator would conduct regular inspection rounds. The positive results with JOIN? A higher safety is assured for the process as well as for the operator. Deviations are detected at an early stage and the risk of downtime is drastically reduced. In an installation with many temperature and pressure gauges, these must be checked regularly, sometimes in unsafe zones. It’s considered a labor-intensive job. The data are now extracted by the Sushi Sensor and sent near-real time to a dashboard where abnormalities can be detected at an early stage. 

There are other examples of JOIN projects. In another installation, oil samples were extracted from hazardous zones and tested in the laboratory. The process is labor intensive and high-risk. Using the JOIN IIoT solution, the oil performance is now tested using sensors. Abnormal values are detected, and oil consumption is greatly reduced because oil is no longer replaced preventively but only in case of deviations. On-site sampling is no longer needed. 
Another JOIN project provided a solution for a situation where a lot of corrosion occurred under isolation. In this installation, this was a problem for pipes, manholes and vessels. Using the guide wave-radar measurement principle and embedded sensors, moisture and corrosion are detected, and the problem locations can be defined very specifically. A switch was made from risk driven inspection to predictive and data driven programs resulting in higher safety and reliability. 

On installation with several manually operated valves, no valve position indication was provided. Using an IIoT solution, the exact position of the valve is now displayed in a dashboard and a signal follows in case of unauthorized or incorrect operation.