The concept of the connected car can mean different things to different people. Some may think about the consumer side where we have services that connect our cars to the manufacturer and allow us to track things like oil use, fuel economy, service reminders, and more through apps on our phones. This helps maintain the service on our cars. For others, connected car might mean being connected via WIFI and having access to telematics data, such as traffic and weather warnings, or real-time access to internet searches to find services and then transfer the location directly into an on-board navigation system. Then to the more techie types, like me, it could mean vehicles that are interconnected to one another and self-aware while traveling the same roads – so self-aware that the vehicles can help to regulate traffic congestion by keeping more consistent speeds or avoid lane change collisions by preventing drivers of side-by-side cars from changing lanes until one car has passed the other. Regardless of what the connected car means to you, it is safe to say that vehicles today have become mobile computing powerhouses with the capability to send and receive mass amounts of data that can be used in many ways. So, how does the connected car relate to the industrial Internet of Things (IoT) and can we make use of the data from those two worlds? Let’s explore the possibilities.
When you think industrial IoT you are probably thinking more about the shop floor where you have machine assets, sensors, and other devices that are connected to a cloud-based ecosystem. This ecosystem allows you to capture the mass amounts of data, coming from the sensors in your assets, into a common data lake. From there you can couple your industrial IoT data with other data sources and analyze it to look for trends, find anomalies, figure out corrective actions, and make better business decisions. Some of the more common use cases are condition monitoring and predictive maintenance. The industrial IoT brings new possibilities to these areas through the ability to do things like remote condition monitoring or remote service. Going even further you can start to look at data not just from one work cell, line, or plant, but all your work cells, lines, and plants together as if they were one. Now you can develop a more complete view of your entire manufacturing ecosystem and look at the interdependencies from one line to another or one plant to another. Sounds cool, right? This is where things start to get even more interesting.
Let’s think of ourselves an automotive OEM and imagine a scenario where an anomaly is detected in the plant where our most common engine is produced. That engine is used in multiple vehicle platforms where the powertrain and final assembly are taking place in different plants in different parts of the world. Through the power of the industrial IoT, we can quickly detect the anomaly and begin to figure out the best course of action to address it. For example, using the industrial IoT data together with our digital twin of production we could start adjusting production throughput in our other plants to balance and accommodate for the anomaly. While in parallel, we can be running more analysis to find the root cause of the anomaly.
This can be done by taking the industrial IoT data that we’ve captured and feed that not only into the digital twin of production but back upstream to the digital twin of the product as well. There we can do further simulations, using real-world data captured directly from the shop floor, and figure out if the anomaly is really within our production system or is it the result of a complicated part design that is creating challenges that we’d not predicted from the original baseline simulations of our production systems. Sometimes the corrective action to address a production anomaly isn’t to change the production system itself – as that can be quite costly depending on the nature of the situation. Sometimes it’s better to go back to the original product design and optimize it further for manufacturability so that we can produce it more efficiently using our existing production systems. Regardless of how you decide to use the data from your industrial IoT solution and marry that with the data from your other systems – none of it would be possible without the concept of a fully closed-loop digital twin, where you can effortlessly bring the digital twin of performance together with the digital twin of production and product.
Now, let’s add some icing to the cake and imagine we are now producing smart connected cars that feed real-time operational performance data via WIFI connectivity to a cloud-based vehicle data monitoring system. In that same common engine as before, we start getting notifications of unusually high oil consumption as compared to the KPIs that we set up in our monitoring system. Using our business process management (BPM) tools, we can start to correlate those notifications to feedback that we are seeing from our dealer service network as well.
At this stage, it is not clear what’s causing the high oil consumption but given that this is a very common engine installed in a high number of vehicles – it is enough to justify the product line owners to investigate further. As a sophisticated automotive OEM, we have a fully closed-loop digital twin of our entire product, production, and performance ecosystem. So can’t we just marry the data from our vehicle data monitoring system, our dealer service network, and our digital twins together – even though the data lives in different systems – and dig into this further? The short answer is, yes. But how can we go about doing it? One potential route is to use low-code application development to quickly create a connection between our industrial IoT data, our vehicle monitoring data, and our service data. From there we could use lifecycle analytics to run analysis on the vehicle data by VIN and find any correlation between the start of the high oil consumption in the field, the production runs involved in producing the vehicles, and the production runs of the engines showing the high oil consumption.
While running these analyses, we find that there was an anomaly during engine production, and through further correlation back to our dealer service data we find the anomaly may have resulted in improperly installed piston rings. According to our dealer service data, it was noted that the high oil consumption is being attributed to improper ring installation. Now that we know the cause of the problem, we can create the corrective action to address the anomaly for future production, we can also figure out how many engines were affected by correlating back to all the production runs, then tie that data back to all vehicles those engines were installed, and proactively start to notify dealers and owners of the issue plus institute the necessary service recalls before further damage is caused by the high oil consumption. As a bonus, the application development work that we did can now be further used to help us with any future issues and can become a standard part of our overall systems and processes.
I hear some of you saying — “But I get vehicle recall notifications today, isn’t this sort of thing already happening?” Yes and no. The vehicle recalls that you get today are often the result of months or years’ worth of problems that gradually trickle in through dealer service complaints. Eventually, there is enough data to figure out there is a problem and the manufacturer can create a corrective action to address it. But, with real-time operational performance data from the Connected Car coupled together with industrial IoT data and other enterprise data – a manufacturer can find a potential problem in days or weeks, not months or years. More importantly, the manufacturer can also take a much more proactive approach to start corrective action and head off further catastrophic failure before it happens – even if dealers or owners haven’t started reporting anything yet. In this more proactive approach, the manufacturer builds an even better relationship with their owners and further increases their brand reputation and loyalty for the future.
While these scenarios are hypothetical, they are based upon many experiences and feedback from various customers, my personal experiences using data in racing, and some of my thoughts and visions as to what the industrial IoT and the connected car has in store for us. These scenarios may seem out of reach to some but given the rapid advances in technology, the amount of data being generated from production systems and vehicles today – you’d be surprised just how close they are. I encourage you to learn more about Siemens’ industrial IoT as a service and low-code application development platform solutions – and if you do happen to be an automotive OEM or supplier be sure to check out our automotive industry solutions as well.