AI in the Machine Manufacturing: Gearing Up for a Smart Revolution

"Artificial intelligence will fundamentally change machine manufacturing and the aftermarket. Companies that don't act now risk losing touch," warned Dr. Matthias Mensing, Director Data & AI at MARKT-PILOT, in the webinar "The AI Evolution: Powering the Machinery Industry's Future" at the end of March this year. His forecast underlines the urgency for machine manufacturers to deal with the latest AI trends.

• DeepSeek - a new "Sputnik shock"? Current developments and forecasts  
• Open-source models are changing the market
• Specialization and autonomy: AI agents in the service sector
• Democratization of AI  
• Artificial intelligence in the parts business
• Bits and bytes on the hunt for parts: AI use at MARKT-PILOT  
• Further webinars on AI in machine manufacturing are planned  
  

The interest of the participants in the online event was correspondingly high. Almost 100 service managers from well-known machine manufacturers accepted the invitation of MARKT-PILOT. Matthias Mensing not only presented current developments but also ventured a look into the near future of the industry. 

DeepSeek - a New "Sputnik Shock"? Current Developments and Forecasts

At the beginning, he addressed the latest developments on the "AI front" and provocatively asked whether the launch of DeepSeek r1 was a new "Sputnik shock" – alluding to the launch of the first artificial Earth satellite in 1957 by the then Soviet Union.  In any case, according to Mensing, China has proven that attractive alternatives to the established AI technologies of Silicon Valley are possible – at lower development costs and with older hardware. 

"The Chinese startup's new Large Language Model (LLM) has shaken up the tech industry worldwide; not least because it refutes the general hypothesis that further significant advances in generative artificial intelligence can only be achieved through the use of more and more hardware resources, but rather because the art lies in intelligently linking of the existing concepts."  

Mensing explained the potentially disruptive effects of this model on the market: The open-source nature and the significantly lower cost per token would make the release of DeepSeek a game-changer for the industry.

This development also has far-reaching consequences for machine manufacturing. In view of the latest developments and the associated advantages, AI will ultimately prevail in this industry as well. According to a recent study by Bain & Company, machine manufacturing companies worldwide can increase their productivity by 30 to 50 percent using artificial intelligence.¹⁾ Mensing emphasized that there is enormous potential in the aftermarket in particular: "We see that AI-supported predictive maintenance systems can reduce machine downtime by half. This means not only cost savings, but also a significant increase in customer satisfaction."  

Current studies confirm the statements. For example, the use of AI can reduce the average repair time by providing precise diagnostic information and optimizing repair processes (mean time to repair).²⁾ In addition, the AI-supported optimization of production processes leads to higher overall equipment effectiveness (OEE) and thus to better utilization and efficiency of the machines.²⁾ 

Open-Source Models Are Changing the Market 

Another trend highlighted by the Director AI & Data at MARKT-PILOT is the increasing in-house development of AI solutions in companies. "In 2024, the share of standardized, ready-made technologies such as ChatGPT in companies has already fallen compared to the previous year. On the other hand, the number of companies that want to develop AI solutions themselves is increasing, also and precisely because of the availability of open-source models." This development opens new opportunities for machine manufacturers to use tailor-made AI solutions for their specific requirements. 

In this context, Mensing predicted a shift in the AI sector. Open-source models are expected to gain significant market share, while established providers such as OpenAI could lose share. Developments such as DeepSeek would refute previous paradigms on artificial intelligence and question the business models of traditional AI providers, according to the speaker.    

Specifically:  

• AI is complex and time-consuming, which is why it is expensive and must therefore be offered at premium prices. 
• Higher AI performance is only possible through more powerful chips (hardware).
• Scaling AI can only be achieved through large data centers.

Specialization and Autonomy: AI Agents in the Service Sector 

Another important trend is the specialization of AI providers, as Mensing explained. The market is moving away from horizontal, all-encompassing solutions to vertical, specialized providers for specific tasks or industries. For machine manufacturing, this means that AI solutions will increasingly be available in the future that are specifically tailored to the needs of the industry. 

According to Mensing, the expected breakthrough of AI agents is particularly exciting for the aftermarket. These autonomous systems, which can make decisions independently, will be used in various areas. "Imagine an AI agent continuously monitoring the condition of a machine, automatically ordering parts when they are needed, and even scheduling maintenance appointments with the customer. This is not a dream of the future but will become reality in the next few years." 

Democratization of AI  

The fact is that the effects of these developments on machine manufacturing and its aftermarket are manifold. The increasing availability of powerful and cost-effective, innovative AI models will also enable small and medium-sized companies to implement corresponding solutions. This leads to a "democratization" of AI technology in the industry, as Mensing emphasized. In this context, the manager spoke of a "revolution in customer care" that AI-supported systems would trigger. Chatbots and virtual assistants based on specialized language models automatically process a large part of customer inquiries, thus relieving service employees of routine tasks.  

New opportunities are also opening in the field of predictive maintenance. Typical scenario: AI systems analyze sensor data in real time and detect potential problems at an early stage before they occur. This enables more efficient planning of maintenance work and reduces unplanned downtime. According to current information, the use of AI in predictive maintenance can extend the service life of machines and systems by an average of 20 percent.³⁾ This is made possible using condition monitoring and machine learning, which allows it possible to accurately predict wear and optimally plan maintenance work. 

A McKinsey study shows that companies that use AI-powered predictive maintenance solutions can reduce their maintenance costs by up to 25 percent.³⁾ This cost reduction is mainly due to the avoidance of expensive downtime and the reduction of component wear. For example, a renowned steel mill uses data analytics based on AI algorithms to optimize the use of rollers, resulting in an 18-month increase in service life.³⁾ 

These figures highlight the significant potential of AI-powered predictive maintenance to improve machine efficiency and service life in industrial applications.  

Artificial Intelligence in the Parts Business  

In the further course of the webinar, Mensing explained the importance of artificial intelligence especially for the parts business. The industry is facing a few challenges that can be addressed using AI. 

This means: 

• Market fluctuations and rising customer expectations: Demand for spare parts varies widely across global markets, while customers expect fast availability and accurate service.  

• Increasing cost pressure: New and international competitors are aggravating price pressure in the parts business.   

• Increasing complexity: Thousands of items, multiple geographies and customer segments, and fluctuating demand increase the complexity of portfolio management.  

• Pricing and competition: Different price points in different regions and distribution channels make it difficult to adopt a consistent pricing strategy, while competitors are quickly shifting to dynamic pricing strategies.    

• Operational and inventory inefficiencies: Excessive inventory ties up capital, while supply shortages can lead to lost sales.  

• Fragmented data sources: The increasing digital transformation is leading to a variety of data sources, such as IoT-connected devices, ERP systems, and e-commerce platforms. 

Matthias Mensing: "AI-supported systems can help to cope with the increased requirements without increasing costs. They enable companies to make better decisions based on greater market transparency and create additional service offerings through automation and the use of digital services." For example, AI-supported forecasting models could significantly improve demand forecasting for spare parts. This leads to a reduction in inventories while increasing availability at the same time, as Mensing explained.  

Bits and Bytes on the Hunt for Parts: AI Use at MARKT-PILOT 

To illustrate the potential of artificial intelligence in the spare parts business, Matthias Mensing presented two concrete cases from the practice of MARKT-PILOT: 

Automating mystery shopping: AI agents automatically query part data, manage email communication, and extract relevant data.

The system formulates initial e-mails, processes incoming correspondence and autonomously decides on the type of response in the context of the existing data.   

Identification of trademark infringements: AI systems automatically detect whether the products offered in web shops are original parts, compatible spare parts or replicas. By using text analysis algorithms and large language (LLM) models with multiple specialized agents, each result can be classified into one of the three categories. 

These examples illustrate how AI not only optimizes operational processes but also supports strategic decisions in the spare parts business. "The integration of such AI solutions will be crucial for success in the aftermarket in the coming years," Mensing summed up. 

Another aspect he emphasized is the possibility of using artificial intelligence to develop new business models in the service sector. "We are seeing a trend towards 'equipment-as-a-service' models, where machine manufacturers no longer just sell products, but guarantee availability and performance. AI plays a key role here by enabling continuous monitoring and optimization of the plants." 

Further Webinars on AI in Machine Manufacturing Are Planned 

Finally, Mensing appealed to the participants to actively engage with the new AI technologies: "The question is no longer whether artificial intelligence will find its way into machine manufacturing and the service business, but how quickly companies can adapt these technologies and use them for themselves." 

The webinar was the prelude to a series of online events planned by MARKT-PILOT for this year on the various aspects of artificial intelligence in machine manufacturing. The following webinars will focus on specific use cases in the industry's service and spare parts business. 

(Sources)

1. https://www.springerprofessional.de/kuenstliche-intelligenz/maschinen/innovationsfaehigkeit-entscheidend-fuer-mehr-produktivitaet/27010934 
2. https://www.markt-pilot.com/en/ai-machine-manufacturing
3. https://ratgeber.volz-witten.de/predictive-maintenance/