Artificial intelligence (AI) is set to be pivotal to supply chains going forward, however, there are several obstacles supply chain managers will have to navigate if they’re successfully going to utilize it.
The state of artificial intelligence in supply chains is still something that is very much unfolding. Both AI and supply chains are multifaceted and thus have elements where they work well together and elements where they don’t. For example, a subset of AI like machine learning has become a prominent feature in forecasting. However, AI being used for self-driving trucks is still years away. Regardless, AI is undoubtedly deeply intertwined with supply chains. It has to be carefully integrated into them in order to work well.
This week’s article by Morai Logistics explores 5 barriers that supply chains face when attempting to make us of artificial intelligence.
Lack of Data
AI only works optimally if it has access to large amounts of accurate data. If there isn’t enough data or the data is of a low quality, the results it produces will suffer. Take machine learning for instance, in order to make predictions or employ its algorithms, a computing system needs enough clean data to pull from for its predictions to be accurate. Simply put, not having a large pool of consolidated up to date data for an AI is like having a sports car without any fuel.
Segmented Artificial Intelligence
Supply chain managers know that one part of their job is to keep an eye on the big picture. Supply chains may be broken up into many individual processes and procedures, but they come together to make the chain. As such, it’s critical that AI implementation be holistic. Much in the same way AI needs access to data that is clean and plentiful, it also has to have access to data that is uninterrupted. If an AI only has data split into disparate segments of a supply chain to work with, then it will produce commensurately uneven results.
Lack of AI Knowledge in the Workforce
Artificial intelligence is conceptually new. It’s also confusing for many. When introduced into a supply chain, supply chain managers may find that many along their chain having trouble adapting to its functions. This is entirely understandable given the often complex and changing nature of AI. With that being the case, supply chain leaders should be providing training for their workforce. Or, conversely, they can hire new personnel to make up for this knowledge gap.
Poor Understanding of AI Processes
Understanding how to use AI and what it’s there for is different from understanding what it’s specifically doing to produce its results. This leads to the “black box” problem—that being the results themselves being mysterious. If there isn’t a transparent AI operation in place, it will likely produce inexplicable results. These results then have to be accepted on faith.
As an article in Forbes highlighted earlier this year,
Black box solutions are controversial. With a black box solution, planners cannot see into the machine and understand how the forecasting engine is generating the forecast. They must trust the output. AI solutions are more likely to be black box than traditional solutions.
Measuring Success
The metrics used to gage success within a supply chain that adopts AI will change. AI transforms many functions in a supply chain. As a result, the indicators of success need to be adapted to these new supply chain realities. Moreover, the nature of AI and what it can do is in a state of continual development. Meaning measuring how successful it is can require continual adjustments of success markers with each development.
