My childhood was spent listening to songs piped out of an AM transistor radio, selected at the whim of a DJ. Today my 4-year-old grandson merely needs to tell his virtual assistant named Alexa the music he would like to hear. Technology has changed so dramatically in my lifetime that it’s hard to envision a future state when there isn’t the kind of exponential growth we’ve seen since the mid-1990s. 

Arguably the most significant technological advance for American business—the Industrial Internet of Things (IIoT), which holds the promise of unimagined productivity gains and untold benefits to consumers—isn’t assured. That’s the essence of a new study, “Smart Factories: Issues of Information Governance,” produced by the Manufacturing Policy Initiative (MPI) out of Indiana University’s School of Public and Environmental Affairs.

MPI Director Keith Belton’s original purpose in creating this collection of analyses—written by experts from academia and industry, and which my organization had a role in developing—was to determine the key policy solutions that could help spur the growth of smart manufacturing. But after conferring with top manufacturing experts in MAPI’s community of leaders, he realized that smart manufacturing in this country faces some bumps in the road.

What prevents its immediate adoption? First, the technologies necessary to create a thoroughly integrated cyber-physical system are still in the early stages of development, and the expertise to operate and maintain these systems is still being cultivated. In addition, manufacturing investment cycles are long—existing factories will likely take years to adopt new “smart” technologies. Most important is the need for rules around the collection, flow, and analysis of digital information—in effect, information governance through the collective action of public and private organizations. As Dr. Belton puts it, “Get the rules right, and the promise of smart manufacturing will become a reality. Get the rules wrong, and smart manufacturing will never fully materialize. Information governance matters.”

Here are some of the biggest challenges outlined by the authors:

How do you hold AI accountable? AI has come a long way in a short time­—from Deep Blue’s defeat of chess champion Gary Kasparov in 1997, to the first licensed self-driven car in 2012, to the use of generative design as an auxiliary to additive printing in recent years. But significant technical challenges still exist, which can lead to important legal, ethical, and public policy questions. For example, how to assign liability when AI makes a mistake? What kind of government oversight is needed when AI is used for security uses? Machine learning is fast, but it’s far from perfect, and the public and private sectors need to clarify the rules governing smart machines.

Who will develop the technical standards? There is no IIoT without the adoption of global technical standards. After all, smart factories will be built upon the integration of IT and OT, which will require uniformity to allow for the widespread acceptance and use of specific technologies. IIoT is complicated, with hundreds of standards already in existence or in development, and potentially thousands more to come. Considering that rapidly developing technologies often outpace the slower standards development process—and considering the decentralized U.S. standards system, where hundreds of standards developing organizations create standards—the need for public-private coordination becomes all the more apparent.

Do the benefits outweigh the risks? No one doubts the benefits of smart manufacturing. And yet developing an integrated IT and OT system that is hooked up to the internet is like leaving your house keys resting on your front porch. Once companies take the plunge, rogue nations, criminal organizations, and hacktivists all potentially have access to their factory floor and supply chains.  Thus, companies spend more and more time and money on cybersecurity risk management, including the training of employees, building security into devices at the start and monitoring outside service providers. There will continue to be uncertainty until a collective framework can be developed that allows for governments, companies and experts to coordinate their approaches to cyber risks.   

As Dr. Belton concludes, collective action is needed to create an environment conducive to investment in smart manufacturing. Manufacturers working with third-party providers can initiate some solutions, but in most areas, only coordinated public-private action will provide the certainty that drives investment.  

Stephen Gold is president and CEO, MAPI (Manufacturers Alliance for Productivity and Innovation).

from industryweek.com