The underlying technologies of Artificial Intelligence (AI) have seen valuable growth over the last few years. Some of its technologies like neuromorphic computing, real time emotion analytics, thought controlled gaming and autonomous surgical robotics are already revolutionizing various industries in terms of precision surgery, real time gaming environments etc. Though these features are making AI very useful, we are still struggling to implement basic cognitive capabilities in AI using adaptive intelligence. AI not only has strengths but also shortcomings when it comes to the vision of replacing humans.
Let us consider application of AI in various dimensions of supply chain management. In case of procurement, we can use a chatbot with Natural Language Processing capabilities to negotiate and solve operational problems. These problems will be solved based on historical data for situational decision making using predictive analysis. In case the condition is new, the problem gets transferred from a chatbot to a human. Similarly, Machine Learning could help in better optimization of inventory, and demand and supply planning based on pattern recognition. Still there must be a human interface to understand and monitor the overall operations planning. Other enablers of AI like IoT and robotic automation can be used in warehouse for stacking, retrieval and order picking under the surveillance of a warehouse manager. Hence, even though AI is replacing us for major operational works, it still needs to be under the supervision of humans.
Now, let us look at the scenario of complete AI controlled supply chain. In this, we are looking beyond augmentation and automation-enabled capabilities. For example, Machine Learning algorithms for complete procurement management, driverless vehicles for inbound logistics, automated production lines and sensor-based robotic warehouse management. The future looks promising for AI-based dynamic supply chain processes. The question for now is, when will this be possible?
As of now, there are various challenges to it. Access to real time data is one of them, as decision making on old data can result in sub-optimal processes. Similarly, getting access to data, which has the ability to affect the business but lie outside the enterprise is also a huge deterrent. AI should also help in executing the decisions rather than just providing backend predictive analysis. The execution here becomes a challenge, as the system includes various partners and stakeholders even outside the organization.
Therefore, as of now, AI is at a narrow intelligence stage and we should look at AI not as a replacement but more as - an assistant to humans. It has helped us in avoiding repetitive and routine works. Going forward, its coordination with humans will be more innovative rather than simply augmenting and automating. Hence, pragmatism demands that we use AI to counter the weaknesses of humans and vice versa. This will lead to huge optimization and value generation with more agile decision-making across all the functions of any industry.