Wisconsin Public Radio

Manufacturing is still critical to the economy United States. Clyde Prestowitz, says it's time to start realizing the positive spillovers that manufacturing creates... Read more  

Events & Activities

Stephen Olson at Chinese Development Institute Conference


 Clyde Prestowitz giving presentation to CDI...


Steve Olson teaching trade negotiations at the Mekong Institute...


Stephen Olson to speak at upcoming workshop organized by the International Institute for Trade and Development on 

"Economics of GMS Agricultural trade in goods and services towards the world market"

Chiangmai, Thailand Sep 8-12.

“Knowledge Synthesis” and Changes in Value Chains for Fully Digital Manufacturers that Use AI and ML

Click Here to Download the Full Report as a PDF File

By Robert B. Cohen, PhD, Economic Strategy Institute
April 16, 2020


This study of “fully digital” firms finds that such firms change their value chains in two phases. In phase 1, they build smart factories and connect nearly all manufacturing processes to machine learning models and AI that analyze machine data rapidly.1 In Sanofi’s new Continuous Biologics Manufacturing Facility, or digitally enabled, continuous manufacturing facility, data from processes is evaluated in real-time and sent to workers’ iPads and displayed on digital screens. This lets Sanofi’s employees act quickly to adjust any processes that need to be changed.

Initially, these firms build a high-performance wide area network that connects all the equipment in a plant; this network connects with data analysis systems. With these systems, machine learning and AI are applied to important processes in the plant. This study finds that moving to automation and analytics has substantial benefits, increasing output and lowering costs.

In the initial digitization phase, firms transform complex parts of their value chain, such as capital-intensive operations with the highest R&D budgets. In autos, this includes assembly as well as the press shop in assembly operations, as well as engine and drivetrain production (two of the main subsystems for auto assembly).

After firms complete this initial phase, they: 1) make more efficient use of capital in optimized processes; and 2) continuously update and redesign the machine learning models that contribute to using AI to analyze processes. This helps firms add additional factors to improve their analysis of processes and operations.
In phase 2, “fully digital” firms take advantage of network effects that have even larger benefits inside firms and for supply chains. In this phase, firms begin to understand that a fully digital ecosystem can take advantage of network effects that have the potential for unprecedented impacts on how they reshape their value chains.

When firms use “knowledge synthesis”, as I have called it, they employ insights obtained by optimizing one machine and create a template or knowledge platform that they can employ to optimize other machines either inside the same factory or in other factories.

In sum, when a firm adopts a “fully digital” infrastructure, the process includes three steps. First, deploying a digital networking infrastructure that links together all machines, underpinned by rapid networks. Second, analyzing the performance of equipment and learning to optimize and change related operations. Third, constructing an ecosystem to apply the insights from the analytics. These evaluate and optimize machines’ performance within a plant and across plants.

Join our mailing list

Latest Publications

The Betrayal of American Prosperity.

The Trans-Paific Partnership and Japan.

Making the Mexian Miracle.

Industrial Policy and Rebalancing in the US and China.

The Evolving Role of China in International Institutions.


Contact us

Economic Strategy Institute

1730 Rhode Island Avenue, NW, Suite 414 |  Washington DC  |  20036
Ph (202) 213-7051  |  Fax (202) 965-1104  |  info@econstrat.org