With its headquarters located in Taiwan, Foxconn, formally known as Hon Hai Precision Industry Co., Ltd., is a worldwide contract manufacturer of electronics. It was established in 1974 and is well known for manufacturing consumer electronics for big businesses like Amazon, Microsoft, and Apple.
It generated a projected yearly revenue of $4.1 billion and employed over 90,221 people worldwide as of 2023. With sizable sites in China and other places, the company’s production activities are spread throughout several nations.
AI has been accepted by Foxconn as a fundamental component of its long-term strategy. Through automation and smart factories, the company is using AI to increase industrial efficiency.
In order to concentrate on smart manufacturing, industrial internet solutions, and cutting-edge technologies like artificial intelligence (AI), cloud computing, and robotics, the company established Foxconn Industrial Internet (FII) as a subsidiary. According to FII, investing in AI-related products, like AI servers, allowed the company to increase its revenue by 200% in the first three quarters of 2024 compared to the same period the year before.
Two interesting AI use cases at Foxconn are examined in this article:
Using digital twins to increase profits: utilizing digital twin technology and algorithm-driven automation to enhance productivity, reduce energy use, and boost profitability.
AI-driven quality control to increase productivity: utilizing deep learning algorithms to improve quality inspection accuracy, decrease errors, and save rework expenses.
Using Digital Twins To Increase Profits
Due to severe cost pressures, growing labor and material expenses, and the requirement to maintain competitive pricing for customers, the electronics manufacturing business is generally acknowledged as a low-margin industry. For instance, despite sequential sales increase of 6.64% during the same period, the Consumer Electronics Industry reported an average net margin of just 13.4% in Q4 2024, indicating the sector’s narrow profitability margins.
Furthermore, Foxconn’s foray into the production of AI servers and electric vehicles (EVs) presents challenges including controlling excessive energy usage, guaranteeing product quality, and expanding operations to satisfy rising demand. Its capacity to sustain operational agility and profitability is further complicated by supply chain interruptions, geopolitical conflicts, and sustainability constraints.
These issues call for cutting-edge technical solutions to streamline operations, save waste, and improve scalability, paving the way for Foxconn and Siemens to work together to construct the “factory of the future.”
Foxconn and Siemens inked a Memorandum of Understanding in 2024 to promote manufacturing sustainability and digital transformation. Integrating AI to power the “factory of the future” is the main goal of their collaboration.
The partnership intends to use the following AI-driven advancements to create a smart manufacturing ecosystem:
Siemens Xcelerator: A package of software and machine learning-driven solutions designed to optimize Foxconn’s manufacturing and engineering processes.
Digital twin technology: Before Foxconn’s factories are physically implemented, Siemens’ digital twin technology will generate virtual versions of them, allowing for thorough simulations and optimizations.
AI for Automation: Putting AI-powered automation solutions into practice to improve manufacturing processes’ accuracy and efficiency, such as contract design and manufacturing services (CDMS) for electric vehicles and electronics manufacturing services (EMS).
Data-Driven Optimization: The process of analyzing massive amounts of data produced by manufacturing processes using algorithms to find trends, forecast equipment breakdowns, and improve production schedules.
We may infer the expected process based on industry best practices and the stated aims, even though the news release mentioned above does not specify the precise AI techniques and implementations:
Data Acquisition: Gather information from sensors, equipment logs, and production records, among other sources, throughout Foxconn’s manufacturing plants.
AI Model Training: To teach AI models to identify patterns and forecast results, apply machine learning algorithms to the gathered data.
The following could be included in the training process:
Predictive maintenance involves using sensor data and past maintenance records to train models that forecast equipment faults.
Process optimization is the use of AI to determine the ideal process variables, such as temperature, pressure, and speed, for different manufacturing stages.
Quality control: Putting in place visual inspection tools driven by AI to find flaws in real time.
Digital Twin Simulation: AI models are used to simulate various situations and build virtual representations of factories using digital twin technology.
Real-Time Optimization: Use AI models to optimize production procedures in real time by modifying parameters according to anticipated results and existing conditions.
Feedback Loop: To increase the precision and efficacy of AI models, continuously assess their performance and update them with fresh data.
illustrating how manufacturers can model and improve manufacturing processes using virtual models of CNC machines prior to actual implementation:
The collaboration between Foxconn and Siemens has the potential to greatly improve Foxconn’s AI-driven manufacturing procedures, especially in the creation of AI servers, which are now a key source of the company’s growth.
AI servers accounted for 26% of Foxconn’s record quarterly revenue of T$2.13 trillion (~USD$64.72 billion) in 2024; this number is predicted to shortly overtake consumer electronics.
The company’s expansion is in line with a strategy move toward AI infrastructure, which is bolstered by Siemens’ digital twin technology and Xcelerator portfolio, which streamline processes and boost industrial productivity.
An NVIDIA blog post about Foxconn and Siemens’ collaboration claims that Foxconn might increase profits in its AI server segment by using Siemens’ cutting-edge automation technologies to eliminate emissions, improve production processes, and reduce energy usage by more than 30%.
AI-Powered Quality Control To Increase Productivity
Manufacturers usually struggle with quality control issues that negatively impact their operational effectiveness and financial success. High error rates in manual inspections might result in more faults and unhappy customers, according to a study released by Nnamdi Azikiwe University’s Department of Industrial and Production Engineering.
Rework and scrap also have a significant financial impact. Poor quality can result in rework and scrap, which can be quite expensive. This emphasizes how important it is to control quality effectively.
A Huawei case study claims that Foxconn and Huawei collaborated to implement Huawei’s Ascend Smart Manufacturing Solution for AI-powered quality control of its intelligent photovoltaic controllers. Foxconn uses artificial intelligence (AI) and algorithms to determine whether silicone grease applied to smart PV controllers is the right color, whether there is enough silicone or not, and whether nameplates are missing, upside down, or attached incorrectly.
The Industrial AI-Powered Quality Inspection Solution from Huawei says it will solve:
- Operation compliance identification
- Identification of defects
- Alignment of position
- Challenges with measurement
According to the study documents mentioned above, Huawei developed a “Industrial AI-powered quality Inspection Solution” that combines big data, cloud computing, and artificial intelligence (AI) for more intelligent quality control by utilizing its experience as a contemporary manufacturing company.
Press announcements from the business state that the platform is based on insights from more than 200 production lines, even though Huawei has not revealed how the solutions operate. More than 800 industrial-grade image processing tools are available in the solution, allowing for accurate fault detection in sectors like electronics and the automobile industry. Automating inspections gives producers a distinct advantage by improving accuracy, lowering errors, and streamlining production.