OK, this week it’s going to be a bit geeky here, but when I heard about the latest FSD Beta release from Tesla and what it does to the autonomous drive systematics, it immediately reminded me to a strategy meeting back in 2015 when introducing Industry 4.0 as the next strategic goal in factory management. But first things first.
Tesla has recently released a Beta version of the new FSD (Full Self Drive) to a couple of testers in the US. When seeing and hearing the first statements about it, it will be a huge step forward towards the ultimate goal, driving autonomously on so called Level 5. As I am driving a Tesla by myself, I already use the FSD a lot on the highway or when driving outside cities. Inside cities I must admit that the reaction time of the sensors and the time that is left for me to make corrections if needed, is making me more nervous than it could possibly easy my time commuting. So, when inside cities or villages I manage driving by myself and that’s OK for now. The reason for that, and that applies probably to all assistant systems in current car fleets, is that sensors measure constantly the situation around the car and then correct immediately when needed, like the ABS system on wet streets. That’s a great achievement and has avoided a lot of accidents for sure. But there are a couple of downsides here – all the cars react on what they sense at this moment in time. Very little of that will be recorded and very little of that will be processed, transmitted, and used for further learnings. The connection of the camera or radar sensors and the ABS system, together with the weather forecast aso. will be missed forever.
Now the new FSD as I understand it and as it is explained in the videos: All cameras, all near and far radar sensors, all the other car sensors and all outside available information together build a second life, a digital twin of what’s actually going on around the car. And then, the FSD simulates the Tesla into this model, as if putting yourself into a video game. In real-time, the car determines relationships and motion of everything in the near surrounding, puts this into context to the desired trip that I want to drive, and then gives the right orders back to the real car to drive me safe and sound. That’s incredibly smart.
Now think about us, the fashion industry, the second polluting industry in the world with our current, quite disconnected business mechanics…
In 2015, we had a strategy meeting about transforming the Hugo Boss Izmir factory from being operational excellent only, into a real smart factory. One of the most important issues was, that we had little real-time information about what was going on on the shop floor and how to manage tomorrow’s workload more efficiently. And as regular cars do with their assistant systems, we used all the real-time sensors and KPIs we had to manage our orders the most efficient way. I remember very well when I talked the first time about building a digital twin of the factory - managing basically two factories at the same time, the digital one just a bit more in the future, the analogue one to follow after. It was hard to understand, blurry and vague. After all, we build an astounding digital copy of the factory and although we didn’t achieve all the desired and bold outcomes, managing complexity without losing efficiency, is one example that I am still proud of. Optimally using the skills of the workforce and integrating real-time information into the management is another one. Looking fearlessly into the coming weeks because you have made lots of simulations of it and know which version you will execute.
The idea of building a digital twin factory or a twin company, like a video game, and then simulate all the issues and hiccups that happen when so many people, machines, products, and processes come together, is a big bet. For me, it has huge potential to make better use of resources and at the same time, as a factory or as a brand, stay or become more profitable. The world is getting ever more complex and normal planning tools and reactive systematics will not make our industry more sustainable. We still produce too many products that end on landfill because of no market needs or wrong predictions. Because of not connecting all relevant information and simulating the best route.
Now, you may say that this is an elephant project and cannot be solved by fashion brands and factories alone, and you’re right. I am happy to see that more and more software companies tackle this - not just singular problem solving, but more holistic approaches. I am sure that more brand will dive into the digital twin mindset and simulate more of their business to make better, more sustainable, and more profitable decisions. Like the new FSD will do when released to my Tesla in a bit - I can’t wait to use it
Enjoy the week…
Comments