With the increasing proliferation of data, connectivity, and processing power at the edge, the industrial Internet of Things (IIoT) is becoming more accessible. However, successful adoption remains out of reach for many: two of every three companies piloting digital manufacturing solutions fail to move into large-scale rollout. Why is it that, despite enthusiasm for this transformation to a digital manufacturing future, few companies have realized its potential at scale?
We already know that AI and IoT at the edge are key to the acceleration of factory transformation, but what is required to catalyze more rapid adoption of these technologies and avoid the pitfalls of pilot purgatory?
In the past two years, we have embarked on a study of over 400 participants across the industry and ecosystem companies—engaging manufacturing leaders and workers, as well as the technologists that develop the solutions and services that support them—to answer this very question and uncover the essential ingredients of industry 4.0. In 2018, we released Phase One of the study, identifying key issues that manufacturing leaders and factory workers are grappling with as they evolve together on the path to the intelligent factory future.
We’ve just released the next phase of this work, Accelerate Industrial, looking at how workers will adopt and react to AI in manufacturing roles—and what strategies and tactics will “accelerate the accelerators”. To date, this phased study represents the most comprehensive view of digital transformation happening in the manufacturing sector.
All Phase Two participants were required to have a first-hand role in a smart factory or a company that develops smart technologies, solutions or services, encompassing the full spectrum of points of view across development, deployment, and maintenance of the technologies within those four walls.
Our research found that while there is a big appetite for digital transformation—83% of companies say they plan to make investments in smart factory technologies in the next two to three years—the people who are the most likely to drive that change are frequently uncertain about how to move forward or hesitant to risk it. So, what accounts for this failure to launch or failure to scale? And how should leaders shift the cultural mindset within their organizations to reap the benefits of industrial IOT?
Here are the top five challenges, cited by respondents, that have the potential to derail investments in smart solutions in the future—and tips for avoiding the perils of pilot purgatory:
Challenge #1: Technical Skills Gap
36% cite a “technical skill gap” that prevents them from benefiting from their investment.
To successfully implement new technology and maintain operations, a company must have a workforce that possesses “digital dexterity”—the people must understand both the manufacturing processes and the digital tools that support those processes.
Solution:
- Create programs that support lifelong learning among the existing workforce, that combine new concepts with hands-on opportunities to use them in the context of manufacturing operations; build modules that are linked so that employees develop and hone their skills over time as they become proficient
- Offer instruction in digital tools and skills (considered important today, but critical for the future). Be holistic in the content by including cybersecurity, infrastructure, AI, data, storage and compute needs. Present individual concepts and their interdependencies.
- Emphasize problem assessment and problem solving before solution implementation
- When standing up a new smart technology project, balance hiring external experts and internal staff to grow your company’s digital dexterity
Challenge #2: Data Sensitivity
27% cite “data sensitivity” from increasing concerns over data and IP privacy, ownership, and management.
To successfully implement an AI algorithm, for example, requires that there is data to train it and test it. This means that data must be shared, yet many companies are loathe to share their data with third-party solution developers. There is also a strong belief that our current data governance policies for internal use within the organization are inadequate to support cross-organizational data sharing.
Solution:
- Formalize data sharing policies for within-organizational data transfer and between-organizational data transfer
- Establish data governance policies that reflect the value of sharing data with the potential risk exposure. Understand that a one-size-fits-all policy will not suffice. Embed customized policies in future supplier/vendor contracts
- Consider data sharing needs prior to standing up a smart project and build in time to negotiate these needs into project operations
Challenge #3: Interoperability
23% say a lack of interoperability between protocols, components, products, and systems.
This is an ongoing struggle that is not new. Today, however, companies are becoming more frustrated with this interoperability as it limits their ability to innovate. It also limits their ability to upgrade system components, since they cannot easily “swap out” one vendor for another or one part of the system for another.
Solution:
- Aggressively pursue and support standards development to increase interoperability; whenever possible participate in consortia such as the Open Process Automation Forum.
- Demand that their vendors work closely together to develop and implement solutions that stress modularity and that offer paths to upgrades over time using multiple vendor solutions.
- Consider open source options when standing up smart technology projects.
Challenge #4: Security
22% cite security threats, both in terms of current and emerging vulnerabilities in the factory.
The combination of physical and digital systems in a smart factory makes real-time interoperability possible—but it comes with the risk of an expanded attack surface. With numerous machines and devices connected to single or multiple networks in the smart factory, vulnerabilities in any one of those pieces of equipment could open up the system to attack. Companies will need to anticipate both enterprise system vulnerabilities and machine level operational vulnerabilities. Companies are underprepared to deal with these security threats, and many rely on their technology and solution providers to do this.
Solution
- Combine OT and IT professionals in smart project teams to assess possible vulnerabilities. Identify people, process, machine and network threats
- Understand the upgrades vendors introduce into equipment and/or operations and anticipate possible changes to vulnerabilities
- Develop “corner case” analyses where no one alternative or feature may be a critical vulnerability but where the interdependencies between alternatives and/or features introduce or increase vulnerabilities. Plan for these nonobvious cases.
Challenge #5: Handling Data Growth
18% cite handling data growth in amount and velocity as well as sense-making.
As AI usage expands, companies will be faced with more data, being generated at a faster pace, and in multiple formats. AI algorithms need to be easier to comprehend—i.e., how does the algorithm arrive at a recommendation?—and these algorithms must be able to combine data that is often of different types and timeframes.
Solution:
- Understand the data that yields business value insights and balance computing at the asset level; bandwidth; and the need for real-time (low latency) control feedback.
- Anticipate sampling rates that reflect changes to machine or operational status. Collecting everything may not make sense.
- Develop a robust system architecture before implementation that balances compute needs and the location of those needs (edge versus cloud, for example), storage requirements today and into the future, and communications infrastructure.
Irene Petrick is senior director of the Industrial Innovation Internet of Things Group at Intel. Faith McCreary is principal engineer in User Experience at Intel.
