Artificial intelligence and machine learning clearly extend the tremendous value gained from data collected and analyzed through the Internet of Things. However, let’s get realistic. AI and machine learning is a later stage development for a vast majority of businesses. The IoT hype is highly driven by dreamers – the innovators who paint incredibly bright futures. However, in today’s environment, most companies are challenged with managing disparate data systems while others haven’t effectively captured and stored historical data. Both of these challenges can impede advancements in the IoT adoption let alone later stage opportunities offered through AI and machine learning.
Beyond critical data challenges, Joshua Bloom, vice president of data and analytics at GE Digital, also referenced a few other challenges specifically associated with AI and industrial machine learning. These included:
- The physical models (for industrial machine learning) do not easily learn from new data, and not all physical laws have been built into large complex systems.
- Trust and acceptance by users of AI and machine learning models.
The latter of these two challenges is inherently human. I have had a history of experiences in market research where business professionals, mostly in sales, don’t trust what the data is saying. But in these circumstances, since we performed the data analysis, it was easier to justify. In an AI and machine learning situation, especially involving massive data sets, will it be as easy to justify the model?
For now, I think it’s fair to assume a majority of businesses need to tackle their data gathering and storage challenges to generate basic visibility into areas either costly to perform manually or that have never been visible before. This can be conducted without AI and machine learning. Then, as these challenges are overcome, these companies can plan a move into AI and machine learning for next level benefits. Let’s start realistically with the IoT, generate some tremendous upsides and then talk about artificial intelligence and machine learning.
Also published on Medium.