This entry was originally posted on the CloudSource blog on march 9th, 2015
As technology becomes ever more pervasive in our daily lives, the questions of how, when, and why we should be using it becomes critical. Concepts are appearing everywhere, ideas are exchanged and visions are established. One of those is called Industry 4.0, catering specifically to the usage of the internet of things in industries. So, let’s take a minute to review how companies are evolving through a clever use of technologies in their industrial operations.
In no way can give a complete overview here as opportunities are immense. Let’s start with the fact that companies increasingly cooperate in product development, across their supply chain and in their maintenance operations. Then, let’s look at where the Internet of Things can actually help enterprises deliver better products, cheaper and faster while maintaining or improving quality levels and services.
It’s important to include the services side as more and more often, companies should think about integrating products with services. That product/service integration is what ultimately differentiates products from one another.
Let’s start with product development. Sure, we have innovations and brand-new products, but let’s be realistic. Most products brought to market are an evolution of previous models or generations. The succession of smartphone models is a good example of this. Given product adoption and customer preferences, companies want to correct short comings from one version to the next one. So, understanding how a product is used, what works and what does not work, is critical to prepare the next generation. Social media conversations can give you a certain view of the shortcomings but you will also need to understand which features are used, and for what purposes. What functions from your smart phone are you actually using? Would it make sense to develop a cheaper version that only includes those features? The price could be drastically lower while the consumer would be getting exactly what they’re looking for.
So, collecting market research as well as user data and then making it available to the developers would really help them defining the next generation product. But given market concerns about privacy, your data collection approach should be thought through very carefully.
An Industry 4.0 example would then be that you, as part of the product development process, desire user data, but you are not interested in the individual. You will need to demonstrate to customers that the information gathered is anonymous and there is no way for anybody receiving the information, legally or illegally to trace it back to the end-user.
This is one example of how product development can benefit from such technologies, but let me share one more. Today’s cars are tuned for average road conditions. They do not adapt to the environment in which they operate. What if, through sensors in the car, by equipping the roads with sensors and by connecting the cars through the cloud, we could tune the engine, the suspension, speed and other parameters to the conditions in which the car operates at any given moment in time? If you’re driving a cobble road your suspension may be looser than on a smooth German motorbahn for example.
Not only would this reduce consumption, it would also make the car safer and the ride smoother. Again, the data gathered during the use of the car could be sent back to improve the mathematical models that tune the car.
These are just two examples of how the combination of sensors, cloud and big data transform the product development process all together.
The collaborative supply chain
Let’s now move to the supply chain. An average supply chain process for an enterprise includes manufacturers, logistics companies and distributors to produce the product and get it in the end customer’s hands. To smoothen the end-to-end supply chain, understanding what happens at each level is critical. Unfortunately, that understanding often does not exist, so buffer stocks are positioned at strategic points in the supply chain to smoothen the operations in case something happens. But stocks cost money, and ultimately it is the consumer who will pay that cost. In the lean terminology, this is waste. The question is then how we can get rid of the waste. Many things have been tried. Probably the most impressive approach is JIT, just in time. But fundamentally, companies following JIT philosophy did nothing other than push the stocks to their suppliers. In the end-to-end supply chain, this did not change anything.
What companies want is “no surprise”. So, if we want to operate smoothly, without buffer stocks, we need two things: early visibility of “surprises” and flexibility to adapt the flow quickly in reaction of such “surprise”.
As was mentioned above, this requires a clear understanding of every step in the supply chain. Ultimately, availability of supply is the critical component in each step of the process.
There are different ways of getting to know when availability of supply is in danger. Here are a couple examples. Social media can highlight events, although outside the perimeter of the eco-system, which may affect the flow. This could include a strike, an accident, a natural disaster, some political event or anything else we may think of. Enterprises themselves may make the status of their production facilities visible by exposing some information from their process and production control systems. If a product line goes down due to a failure of some machinery, this can be made visible quickly. Logistics providers can have tracking systems highlighting where specific shipments are.
But what is important here is not just making the raw data visible, but also analyzing it and identifying what to do to minimize the disruption. Here is where an eco-system “control tower” plays an important role. Software algorithms can spot the disruption, they can serve as analysis tools to identify what could be done and what the implications would be, but it is up to the human beings to take the final decision and communicate it across the eco-system. Whether that is done by the OEM in the supply chain or by a team of individuals representing the key members of the supply chain depends on the implementation. But one this is for certain: that Industry 4.0 increasingly focuses on collaborative working relationships between the players in the eco-system.
Whether we talk about maintenance operations within the production environment or services to maintain equipment at the customer site, the problem remains the same. When is an intervention required? Typically we have two approaches. Either regular preventive maintenance (for example yearly) or maintenance triggered by usage (typical in the car industry), it always happens before the fact and does not take into account the actual status of the equipment.
In fact, some car manufacturers are trying to take that into account. My car has told me for about 15000 km that I was due for a service, the replacement of my brakes, “in 3000 km.” And for a while, that number actually went up. Apparently, I do not wear my brakes down as much as others do.
To answer some of these issues, I’ve seen companies putting sensors on pumps to understand when they needed replacement rather than replacing them on a regular basis. I’ve also seen companies including sensors and intelligence boxes in their products so their service personal could look at the records remotely and figure out when the equipment needed service, avoiding unnecessary visits. By doing so, they actually provided a better service at a lower cost.
Industry 4.0, it’s there but more to come
I could go on with many more examples across more and more industries. As I mentioned at the beginning of this post, my intention was not to be exhaustive, but rather just to give you some hints on what can be done. The explosion of sensors, the improvements in wireless technologies and the appearance of new mathematical models allow many new applications. Our imagination will be the limit of what we can do. However, there remain some physical constraints we cannot solve. For example there is a limit in the amount of data we can share wirelessly. So, new approaches will need to be developed. Centralizing all information to take decisions is probably not where we are going. Intelligence will be pushed at lower levels so that fast interactions can take place close to the generation point, while consolidation and longer term trending/decision process are taken higher up the architecture. It’s a change from today’s approaches. Let’s see how they change our way of life