The term ‘Smart Factory’ and its relevance to apparel manufacturing industry have been discussed so many times in recent years. Most of these discussions just share concepts and no clear implementations of Industry 4.0 can be seen in apparel factories on a large scale. However, there is one brand which is known for its digitally connected apparel factory in the beautiful Izmir city of Turkey with the extensive use of concepts such as AR/VR, cloud computing, data analytics, AI, IOT and ML – and that’s HUGO BOSS.
It was back in 2015 when HB Izmir started converting its largest production unit – on an area of around 65,000 square metres – into a smart factory which produces suits, jackets, shirts and coats with the help of almost 3,500 workers.
The whole factory works on Digital Twin, Robotics and AI which have led HB Izmir to see massive cost reduction, quick turnaround time, flexible processes, minimal complexities, easy retrieval of data at any given point of time, quality improvement and productivity increase.
Now, it’s important to know how HB Izmir could integrate all these complex technologies in its factory at a time when most of the apparel manufacturers either struggle to do the same or don’t even believe it can be done!
Erkut Ekinci, Head of IT, HB Izmir revealed that the factory was able to transform into smart factory from four points of view – Strategic, Tactical, Technical and Methodology.
“For this to happen, we needed to connect all the four pillars – people, machines, products and processes of our factory altogether and that was possible with Industry 4.0,” informed Erkut.
This digital networking between these four pillars makes it possible to create a ‘digital twin’ of the factory which may sound like a sci-fi movie, but it isn’t. The factory is using more than 1,600 (WiFi-enabled) tablets throughout the location which help the management to track any sort of data in real time.
The factory planned this ‘smart’ conversion process in three phases – 2015-2016; 2017-2018; and 2019-2020. Till now, a number of projects have been successfully completed.
- Smart Data Management (SDM)
Data is the most important parameter to focus in any of the smart factories. HB Izmir considers tracking of data as a key to the whole equation. For over 3,500 workers in the factory, the team of developers has designed data collection machines (tablet) which is called SDM. It can collect and display any information related to production, performance quality in real time. “This is a bi-directional shop floor management tool for us that helps us extensively in follow up of KPIs online,” shared Erkut.
- Group board & Line board
Unlike traditional factories where boards on shop floor are still maintained manually, HB Izmir does it all digitally for everybody in the factory to stay updated in real time. Group board is a board to track the information of a group of around 10-15 workers who work together on a certain project, while Line board is meant to track data of all production lines, warehouses, inventory, etc. All the machines are interconnected with each other and these boards display the efficiency of a single operator, entire line, entire shop floor; quality defects, real-time stock/inventory update, shipment information of products and group productivity. “The board also displays the quality parameters of a particular product which is being produced in a line which is generally done manually through papers in factories,” asserts Erkut.
- Augmented Reality (AR)/Virtual Reality (VR)
We all know how soft trainings and their evaluation are done in apparel factories. HB Izmir goes miles ahead in training programmes of its operators and uses AR and VR concepts for operators’ soft training, quality process integration, product information and machine information.
For example, the factory has developed AR tablets that help operators capture the machine image. Once the image is captured, the tablet uses this virtual image and plays a video process of the maintenance and lubrication process of this machine, so that the operator can understand the whole process by himself, over and over again. This training is also made possible by HB Izmir by scanning machine QR code in the tablet.
When it comes to VR, it also is aimed towards enhancing and digitalising the Product Development Training process. The screen of VR tablet shows a list of products and list of operations in each product. A video then can be played for each of the operations for operators to understand the critical points of a particular product. Once the video is watched by the operator, he/she needs to undergo an examination, the questionnaire of which pops up in tablet itself. If they fail in this exam, they need to re-watch the video until they get perfection in training modules.
- Active Digital Assistance (ADA)
Suppose one of the key members of any department of the factory leaves the organisation and a newcomer takes over the position with no idea at all about where the entire data of products, shipment inventory, etc. is kept…Irregularities, time consumption, process overlapping and non-valuable activities remain a hurdle with such a process.
Now, what if the entire factory data is tracked and retrieved at the fingertips? ADA or live Chatbot in HB Izmir eliminates all the above-mentioned barriers, as it keeps an eye on all information – real-time and historical. A person can ask anything about a particular production line, shop floor, shipped orders, ready-to-ship orders and quality to get the desired information in no time.“This ADA assists in Question-Answer format, you just need to ask something to get an answer,” comments Erkut.
- Speech Recognition
Every time when a new order is placed by a buyer, the design and development team needs to define process flow of it. The concerned persons just look at the sample garments and try to create the model flow (design flow) by data entry of 100s of different operations. Normally this data entry is done manually and that’s an arduous task. “We do it through Speech Recognition technology where an operator speaks about a certain product/operations, the tablet recognises the words spoken by the operator and then it discloses all the information. It saves 2-3 hours in each new SKU while defining the operation flow,” says Erkut.
AI-based projects help HB Izmir improvise
It’s incredible for HB Izmir to set an example with its smart factory where no area goes unnoticed when it comes to data collection. Right from planning to logistics, data plays a key role and the analysis of this data provides information about where there is potential for improvement, or even where risks arise. This enables machines, resources and processes to be managed based on well-founded, digital forecasts. Apart from the completed digital projects (above mentioned), HB Izmir also works on some AI-based projects to see a better process flow.
One of the initiatives is Quality Defect Prediction through AI. The factory is able to study the historical data and predict possible defect in operations and products. It claims 80 per cent of defects can be identified using AI and data analytics. “We collect too many attributes about a product including machines, people and processes to make it happen. Our AI tool looks back at all the defects that have occurred in the past in all SKUs we have worked on, and identifies who was responsible for those defects, in which lines these defects occurred and when these were occurred. This process comes up with a prediction of defects in any ongoing project and helps us to take a call way in advance,” informs Erkut.
Similarly, in pre-sewing area, HB Izmir uses AI and Machine Learning (ML) to ‘Precognise’ the possible defects in parts of its auto-cutters. “It is an early warning system for upcoming incidents on cutters, and this way, the preventive maintenance has become an easy process for us,” shares Erkut.
Another significant improvement done is in the HR side. It’s a fact that applying AI concept to analyse human emotions is not at all easy, but HB Izmir has attained it through its ‘Oxygen’ project. This project helps identify low performers on factory’s shop floor in advance by analysing operators’ historical data and defining a trend of their emotions. Once human emotions are captured, the factory management then lays a plan to improve his/her productivity. “It’s not at all easy to capture human emotions, but we are satisfied with the results,” concludes Erkut.
Also Read: Connected factories or robotic automation: What to choose and why?
