The book “Competing in the age of AI” provides deep insights into the impact of AI on businesses. The authors discuss how companies can use AI to become more competitive, and why data and analytics is crucial to the success of a business. Marco Iansiti and Karim Lakhani give many examples from successful technology companies and what was required for them to become an AI-first company.
Prime examples for such twenty-first century “digital-first” companies are e.g. Ant Financial in the Financial sector, Ocado in grocery delivery or Peloton in the fitness sector. The book details many more case studies from “digital-first” companies such as Uber, Amazon, Microsoft, Google and many more. The authors argue that the current era is far more revolutionary than the Industrial Revolution and predict that the benefits that came with mass producing physical goods will no longer facilitate further growth.
According to the authors, growth at traditional companies will hit a point beyond which the organization will “suffer from diseconomies of scale, scope and learning.” By contrast, organizations that follow a digital business model with algorithm-driven operating processes are “almost infinitely scalable.”
While the efficiencies of mass producing physical goods were based on vertical specialization, new digital-first companies break down what the authors consider to be anachronistic siloed organizational architectures. Instead, digital-first companies build data-centric operating architectures which affect all of their value generating activities and thus scale horizontally across the organization.” As a result, “competitive advantage is shifting away from vertical capabilities toward universal capabilities in data sourcing, processing, analytics and algorithm development”. This is a radical but not surprising conclusion if you have followed the rise and growth of digital-first giants in the last decade. The question remains if and how traditional companies who produce and sell physical goods can compete and still grow in the digital era.
In today's times, digital networks and Artificial intelligence are fundamentally transforming business and society. What is needed for a company to succeed in „the age of AI“? A new kind of business model, with software, data and AI as the foundation of a new operating model:
How can companies succeed in the age of AI if their core business model can not directly be transformed into a digital business model? The answer is: by digitizing and transforming their operating model. Let’s now have a detailed look into the details of how this can be applied to manufacturing companies and what benefits manufacturers can gain from applying AI to their production processes.
The operating model is what the resources and employees of a company actually do every day to generate value. Thus, the operating model is the enabler of a company's value and at the same time limits its growth. There are three elements to consider for manufacturing companies who want to improve their operating model: scale, scope, and learning.
Scale: Managing scale is about creating operational processes that deliver as much value to as many customers as possible at the lowest costs. In manufacturing, improving scale typically includes increasing production volume, e.g. by implementing lean production processes or adding additional production lines with new and better machines. For “digital first” companies who follow a digital business model, scaling is typically about removing humans from processes and replacing them with AI-driven algorithms. For example, in the case of Ant Financial, AI driven software determines whether a loan will be given to someone or not. By using AI, a bank can scale this function in orders of magnitude compared to traditional methods.
Scope: a company's scope is defined by the range of value-adding activities it performs. In manufacturing, this includes the variety of products a company produces. By introducing new production lines or new technologies, a manufacturer typically can extend its scope. “Digital-first” companies usually extend their scope by applying software and generated data from one application area to new domains. Ant Financial extended its Loan Processing service with Ant Fortune, a personal investment service, Zhima Credit, a social credit scoring system, MYBank, an internet banking services and a variety of other services.
Learn: the learning component of an operating model is about continuously improving operational processes. In Manufacturing, Continuous improvement (KAIZEN) is an integral part of Lean production and well established in many companies. For “Digital-first” companies, learning is also crucial and most often part of their DNA. The more users a software service has, the more data it collects. And data is the fuel of AI applications and can be used to continuously improve the service. For example, the quality of recommendations within Netflix directly relates with the number of users and user engagement.
Operating models can take various forms. In some cases, they only manage flows of information (e.g. Ant Financial, Google, Netflix). In other cases, operating models define how physical products are built, delivered or operated (e.g. Amazon, Ocado). In all cases, the transition to a “digital operating model” means that humans are removed from the critical path of value delivery. Instead, humans are moved “to the edge” to oversee and control the value delivery process. It is important to note that AI is not replacing humans, but simply frees them up to do different tasks than before. Humans still add value, but they are no longer the limiting factor for scaling a business.
Successful digital-first companies were built on an integrated, highly modular digital foundation. They are prepared to rapidly scale, extend scope and continuously improve. Information technology forms the operating core of such a company and is not only seen as an enabler for the operating processes.
What does that mean for manufacturing companies? How can the production system of a company be transformed to gain benefits from AI?
Since the beginning of mass production, manufacturing companies benefit from economies of scale. Mass production concepts like automation, specialization and standardization help to generate more output at lower costs. Automated production lines and robots increase productivity. Standardized repeatable work and lean manufacturing concepts remove waste from the core production processes. However, market requirements are constantly changing and production is becoming more and more complex, e.g.
As a result, production complexity is increasing. Traditional operating models and production systems create serious constraints for growth. A new model is needed to flexibly scale production processes.
In their book, Marco Iansiti and Karim Lakhani emphasize that “the deployment of Enterprise IT did not transform the trajectory of operating models...IT systems such as Oracle financials and SAP product lifecycle management improved the performance of many traditional operating processes, but these IT systems generally mirrored the firm's siloed and specialized architecture.” The application of automation technology or SAP systems does not change the structure of an enterprise. In most manufacturing companies, processes and software are still embedded in siloed organizational units. Data is most often collected and processed in an inconsistent fashion.
Traditional organizational models that operate through a variety of specialized and siloed organizational processes will no longer facilitate growth. For the flexible organization of production, a software- and data-driven organization is needed. Systems that unify data across traditional Enterprise IT can act as the single source of truth.
On top of such a Unified Data aggregation layer, an additional layer of software is needed. After all, data alone is not actionable. It is crucial to make data actionable, to gain insights and derive actions. A system like WORKERBASE can act as a process improvement layer to do exactly that.
A horizontal data-driven architecture is the new foundation for flexible operations. On top of that, analytics and AI functions can be applied to generate additional value. For example, the high complexity of production planning processes can be addressed with AI-based planning algorithms. Machine downtimes can be reduced by applying machine learning algorithms for predictive maintenance. Inventory levels can be optimized with AI-based material management and automated guided vehicles.
This all requires change management and software systems which connect human workers to data-driven production systems. Because, again, human workers will not be completely removed from operations, but responsibilities and job types will certainly change. Supervisors will become Coaches, Machine Operators will become problem solvers and Forklift drivers will oversee an armada of automated guided vehicles. With a digital operating model, human workers design and oversee an algorithm-driven digital organization. This removes traditional operational bottlenecks. Of course, this also changes the role of management. The times of micro-controlling employees who perform routine tasks are over. Managers must act as leaders and need to design, control and improve advanced AI-driven digital systems.
A new generation of IT production systems is built upon a data and event-driven architecture. App-based interfaces to such systems allow employees to collect data with their smartphone. Wearable devices such as industrial smartwatches augment human workflows and create new data points. The Industrial Internet of things with connected machines and sensors increases volume, velocity and variety of machine data. AI-driven analytics offers new insights and enables predictive production processes. This is not science fiction. In fact, the WORKERBASE Connected Worker platform is such an data- and event driven software system that connects human workers to the production system of a company. Frontline workers interact with production processes through mobile apps. The system creates real-time data and offers 360 degree visibility about all human workflows, if needed in a fully anonymized way. In addition, the system breaks down organizational silos and facilitates collaboration across functional areas so that colleagues can support and learn from each other. And pluggable AI components allow the optimization of production processes in real-time, thus enabling predictive operations.
However, as Marco Iansiti and Karim Lakhani conclude in their book, many traditional firms are not moving forward but protect their backwards-oriented systems built over decades. “Despite the massive business potential of the data-centric operating architectures driving AI-powered firms, many traditional firms hesitate.” say the authors, “they either do not see their architectural problem or are not willing to fully commit to the organizational transformation that is required to solve it.”
Transformation is hard. Really hard. It is a journey which takes a long time. It changes the way an organization functions, how humans are working, as well as the organization’s culture and value systems. If you are interested in how other leaders have started their digital transformation and managed to digitally enable their employees as a first step, please do not hesitate and contact us. We look forward to hearing your story and exchange opinions!
Author: Thorsten Krüger, Co-Founder WORKERBASE. Contact me at firstname.lastname@example.org