Hype is a crucial component to introducing any emerging technology into the marketplace. It draws attention and, in many instances, entices organizations to come out onto the bleeding edge. However, at some point, manufacturers need to move beyond the hype and realize the return on their investment.
A recent report by Lux Research, Artificial Intelligence: A Framework to Identify Challenges and Guide Successful Outcomes, takes a deeper look at the current state of AI. The goal was to provide companies with an outcome-focused framework to make more successful investment decisions and better manage their AI projects.
When properly selected and utilized, AI has the ability beyond the hype, explains Cole McCollum, lead analyst for this report. “We’re seeing AI substantially improve manufacturers’ operational efficiency and quality control. AI provides both the capability to automate simple, routine tasks and to extract insights from datasets too complex for human understanding,” he says.
According to McCollum, some of the most impactful AI applications include predictive maintenance to foresee impending issues that could cause equipment down-time, enhancing the capabilities of automated inspection systems with computer vision, improving internal knowledge search using natural language processing, and optimizing production with machine learning-based recommendations.
McCollum tells IndustryWeek, the key to success is to prioritize quick wins early in the process. “This means investing in use cases where AI is more mature and considering how the challenges of scaling a solution across an organization such as data availability can be overcome,” he says. “Often, these mature use cases will be narrower in their impact and will augment human capabilities and existing processes rather than attempt to fully automate a task. After generating evidence of successful adoption and a tangible ROI from proof of concepts, companies can then direct the further resources needed to scale to a wider rollout and invest in more transformative AI projects.”
Framework for moving beyond the hype
The Lux report calls out four major factors in making the right AI investments and decisions:
1. Clearly understanding the outcomes implementing AI will provide for their business. “Start with the end in mind by determining the problem and outcome to solve for. Then, work backwards to determine the level of AI needed to solve that problem and whether it is feasible with today’s tools using the AI framework. Following that step, identify the key challenges in development and deployment and determine whether emerging solutions will be capable of mitigating those challenges.”
2. Focusing on an AI product's capabilities instead of flashy marketing. This means understanding how the vendor’s use of the technology translates into addressing a business objective. “Opportunities exist to leverage this capability, whether for scaling basic human pattern recognition capabilities, emulating expert pattern recognition, or uncovering patterns in data too complex for a human to recognize.”
3. Knowing when the technology is mature enough to mitigate risk. “More complex the environment, the more immature the application will be with today’s technologies. Applications that require longer-term planning or reasoning capabilities from a human are likely many years away from full, successful implementations.”
4. Identifying practical challenges to both implementation and maintenance of the technology once it is in place. “Seek to get buy-in from the end users of the application as it is being developed to ensure that the end users will both trust and use its predictions. Likewise, because machine learning is an inherently statistical tool, adoption will be more successful in applications that can tolerate some inaccuracy.”