Industrial revolutions may bring sweeping change, but not necessarily all at once. The first such revolution may have transformed manufacturing, but it took from roughly 1760 to “sometime between 1820 and 1840” to do so, as Wikipedia puts it.
Next in line is “Industrie 4.0,” the German term for the next industrial revolution that follows three historical manufacturing upheavals. The first was based on mechanization and steam power; the second on electricity and mass production, while the third was centered around computing and automation.
Now, technologies such as connected sensors, advanced analytics and digital twins promise to help drive Industrie 4.0. The only problems are that realizing that vision can be arduous. While the vast majority of manufacturing professionals believe in the potential of smart factories, the industry has been slow in realizing their dreams of transforming their operations with IoT and analytics.
At Hannover Messe, Cisco representatives explained there are three primary hurdles on the path to Industrie 4.0: cybersecurity, project complexity and data management difficulties. In an interview, Cisco’s senior director, senior director, manufacturing solutions Bryan Tantzen shared his thoughts on those three hurdles and the company’s plans to help address them.
What do you see happening with cybersecurity in the factory landscape this year?
Tantzen: This year, we are seeing [factory owners] terrified about security because they are in many ways very unprotected, even without IoT. Security has gone from something people knew was important to something they consider urgent.
In the past, factories relied on air-gapping for security. But this year, we have seen factory after factory go down. Some of the attacks like WannaCry took out every Windows device in the facility. It could be a robot controller, a PLC, a maintenance workstation. Once the malware got in, it could go across the factory and could go other factories in some cases.
What’s a high-level overview of Cisco’s approach to improving security on the shop floor and what is new there?
Tantzen: Factory security is still not easy but we are looking to make it easier and deployable at scale
First, there is asset discovery. Factory staff often don’t know what is on the plant floor and what is connected. Being able to discover the OT devices and monitor them is fundamental capability one.
The second is being able to have a managed network linked into identity services with zonal isolation and network segmentation.
The third thing is threat monitoring.
If you have those three things, you can still have your ten-year-old systems that are not themselves secure but still solve the vast majority of the problems. We are focused on building a foundation in the factories.
Cisco also has done a lot of work on connected factory and converged factory-wide Ethernet to connect securely using a flat architecture. We have updated our factory security architectures and have issued a security whitepaper with Rockwell. It is providing visibility so you can see when there is a threat. The paper recommends segmentation and zonal isolation, so if something attacks you, it takes out a factory cell but not a whole factory.
Another thing that is essential to Industrie 4.0 is dynamic access control and enabling secure remote access.
In the past, we worked with the oil and gas sector on cybersecurity. We think this is the year to secure the manufacturing space.
What is your approach to addressing the complexity of Industrie 4.0 deployments?
We have to make things less complex as we bring in things like the capabilities of IT like security, mobility, cloud and manufacturing. We are continuing to move the ball forward on network automation so that it is plug and play and purpose-built for OT but having the capabilities of IT.
What do you make of the data challenges in smart factories?
One of the barriers to Industrie 4.0 is so many people want to help you to get data from factory floor devices but they want it going into their stack. It is monolithic.
To really be successful, you have to set the data free. It is not just monolithic but there are all of these proprietary protocols
We are doing a couple of things in this regard. [The Cisco IoT data platform] Kinetic is addressing complexity, data management and bridging data silos.
What Kinetic is designed to do is to bridge those proprietary protocols and you can now connect deeper into the plant. You can not only connect your PLC controllers but connect to sensors directly, robots, CNC machines, location tags for inventory and connect to worker locations for safety.
We are announcing an interaction with IBM using IBM Watson to make it easier for customers to deploy IoT and analytics at scale
What else is new?
We are launching a package to rapidly connect to CNC machine. Most CNCs haven’t been connected at all. So starting to get that data, we are seeing impacts like 10 percent utilization increase.
In some cases, you will have a three times difference in productivity in the CNC machine from factory A to factory B with the same product workload, and [factory managers] had no idea why. Now, you can get the real-time visibility to see what is really going on.
We have also deployed a production system with Kinetic in the state of Tennessee and are getting real time sensor data on road conditions. They can adjust speed limits and warning signs. Tennessee was an area where you had 54 people either injured or killed in an accident several years ago because of fog. Their new traffic system enables them to take action using relevant data.