MxD’s 22,000 square-foot manufacturing floor

ML Journal

ML Journal

M2030 Enabling the Future of Enterprise-wide AI/ML

Augmenting AI/ML with enterprise systems allows manufacturers to create self-learning knowledge systems that will help them transform their businesses by 2030.  

Manufacturing in 2030 will largely be the story of next level digitization on steroids. The popular adage Every Business Is a Digital Business, will become a living reality for global manufacturers. The pandemic-led changes in customer behavior over recent years will continue to accelerate trends towards software solutions and data platforms. Agility, dynamism, sustainability, and customer-centricity will change business models and organizational structures around concepts of autonomy, Servitization, self-fulfillment, circular-economy, and virtualization. Data-driven Servitization models will become core for manufacturers and digital technologies will help companies fundamentally reinvent internal processes.

The upcoming decade is also the story of a transition from the current Industry 4.0 paradigm, to the Industry 5.0 paradigm, the next step in the evolution of manufacturing processes around the notion of machine-assisted humans. Industry 5.0 will also see the next stage of industrial IoT, as well as the widespread adoption of industrial AI – the systematic combination of AI technology with industrial systems that include humans in the loop. Driven by this human/robot collaboration, organisations will be more empowered to improve their triple bottom line of People, Planet, and Profit.

Today, artificial intelligence is already completely transforming the services sector. Manufacturing is next. Powerful cognitive applications are slated to transform the entire gamut of manufacturing operations over the decade ahead from production to supply chains to aftermarket activities.

“Delivering value will also rapidly shift from products to data-driven service models for most B2B companies in this decade.”


Delivering value will also rapidly shift from products to data-driven service models for most B2B companies in this decade. While the product itself may previously have been the link holding together B2B relations, this decade will be about customer centricity with a key focus on individual customers and their purchase journey from product identification to aftersales. For manufacturers, this trend implies the transformation of core business models, initiating new operating paradigms and the monetizing of data. This makes Artificial Intelligence and Machine learning (AI/ML) a top technology priority over the next few years.

Opportunities and Challenges

A plethora of AI/ML use cases that have already been identified. AI/ML is already becoming recognized as a key technology to optimize everything from product design, production, sales, and supply chain to aftermarket services. Manufactures are now increasingly eager to leverage AI/ML to optimize their most fundamental decisions such as:

  1. Portfolio optimization – What is the right mix of products, suppliers, assets, people, and even customers? What is the most optimized way of allocating resources?
  2. Value optimization – How do companies optimize resources on a day-to-day basis by balancing countervailing outcomes (e.g., inventory levels vs. order-fulfilment rates)? How do they improve their situational awareness and optimize next course of action?
  3. Growth Optimization – How can they better predict future scenarios emanating from today’s strategic decisions and make choices that minimize risk and maximize profits for the future?

The central premise of AI/ML adoption is creating these new decision models, which are essentially closed-loop processes where business rules percolate down, and information rises up, with a focus on optimizing decision-making every step of the way to uncover a global optimum for the organization. AI/ML can then become a core factor in smart manufacturing, smart supply chains, connected customers, product as platform, and connect channel models – developments that are slated to unlock trillions of dollars’ worth of value for manufacturers in the current decade

That begs the questions why is AI/ML still a novelty for most companies? What is inhibiting organizations from scaling AI/ML? And what can be done about it?

“Augmenting AI/ML with the ERP system is the next logical step in the evolution of ERP, from a SSOT to a self-learning organizational knowledge system.”


The most common roadblocks for organizations as they look to scale machine learning across the enterprise are, perhaps not surprisingly, time and complexity. The time required to customize, deploy, and sustain new AI-enabled models is very time consuming. The next challenge is the complexity of data. Both data volumes and quality are critical in scaling AI/ML. Last, but not least, is trust; getting organizations to trust the AI/ML and taking it to the next level by making AI/ML autonomous for all routine decision making.

What organizations need now is to alleviate these roadblocks and target the areas that unlock most value. Only then can AI/ML break free from proof of concept to true enterprise scale.

Augmenting AI/ML with Enterprise Systems

Enterprise Resource Planning (ERP), in conjunction with WMS, MES, and PLM, has a strategic vantage point that provides end-to-end visibility of any organization’s entire operations. The cloud-based modern ERP system’s role as the single source of truth (SSOT) for most organizations is well recognized.  Augmenting AI/ML with the ERP system is the next logical step in the evolution of ERP, from a SSOT to a self-learning organizational knowledge system. That augmented system can optimize a business, from the shop floor to the top floor. The sheer span and reach of the ERP system makes it an obvious first choice for organizations looking to achieve scale and unlock most value. Organizations looking for quick wins and rapid scale can get a glimpse of the enormous value of AI/ML enhanced ERP solutions with these potential use cases.

  • ERP as a Self-Learning Knowledge System: A closed-loop self-learning ERP system that can transform business and manufacturing processes and help optimize decision making.
  • IoT, AI/ML, and ERP Integration: When an ERP, with the integration of AI and ML technologies, has access to IoT data, the system is empowered to bridge intelligence gaps, which many businesses face in pursuing new business models such as Servitization.
  • Improving OEE of Capital Equipment: AI-ML, IoT, and cloud ERP systems serve as an always-learning knowledge system that can monitor equipment data in real-time, recognize failure patterns, and help organizations transition from time-based to a predictive maintenance paradigm for equipment assets.
  • Improving Product Quality: With ML, an ERP system gets tracking /tracing capabilities which can help any business predict which product or process characteristics cause failure and enable close-loop product design based on the product’s entire lifecycle.
  • Integration of all Enterprise Systems: Systems such as CAD, PLM, MES, and CRM sometimes counteract each other due to organizational silos. Using AI and ML to close gaps between PLM, CAD, ERP, and CRM systems, enables more holistic business decision making.
  • Impeccable Customer Service and User Experience: Leveraging AI/ML to channelize customer interactions with users by tracking and resolving customer issues and, more importantly, learning customer behaviors and purchase considerations to help marketing and sales teams become more customer centric.
  • Integrated business planning: AI-enabled ERP also provides the ability to conduct scenario-based planning with an overarching objective of maximizing profit, reducing resources, and minimizing risks.

What do customers want to use AI-ML to solve?
Customer experience and process automation represent the top AI/ML cases

Source: Algorithmia_2021_enterprise_ML_trends.pdf 

Companies aiming for exponential adoption of AI/ML in their organization should focus on AI/ML enabled decision making as a competency rather than focusing on specific tools and technologies. They also need to consider how to best integrate previously siloed ERP products with cloud technologies and new AI/ML functionality to ensure that all their teams, business units, and locations can make decisions in sync with the real-time state of operations. They should also be aware that the wide-scale adoption of AI/ML is most successful when its deployment expands beyond dedicated data-scientists to the wider organization. So, having programs or partners in place to help spread awareness and develop new skills among key team members across the enterprise can help drive faster adoption, encourage front-line innovation, and deliver greater value.

“Companies aiming for exponential adoption of AI/ML in their organization should focus on AI/ML enabled decision making as a competency rather than focusing on specific tools and technologies.”

These considerations are crucial in today’s business context, where customers are keen to leverage AI but hesitate because of complexity, time and cost associated with such projects.

Looking Forward

Today AI/ML finds itself at the same juncture that the idea of process optimization was a couple of decades ago when it was still a novelty. With the widespread adoption of process optimization in recent years, it has blended with the enterprise application landscape. AI/ML solutions are likely to follow the same trajectory and their widespread use will make them indistinguishable from a manufacturing company’s core ERP in the future.

The challenges to successful enterprise-wide AI/ML adoption may seem substantial right now, but the opportunities for additional value for manufacturers are enormous in the decade ahead. A new era of AI-enabled value is now within the manufacturing industry’s grasp.  M

About the author:


Chirag Rathi is a strategy lead for Infor’s Industrial Manufacturing business unit.




View More