There is always a simple starting point from where infinite possibilities arise. In the case of logistics, that was an algorithmic problem simply known as the Travelling Salesman Problem (TSP), i.e., determining the shortest possible route a salesman could take to visit a given list of locations and return to their origin. In other words, how do you send a package in a way that was both timely and fuel-efficient?
At a very basic level, there are ‘n’ number of ways to deliver ‘n’ items. That is, to deliver two items, 1x2 (so, two) combinations are possible. For three, 1x2x3 (six) combinations, and for delivering just six orders, an atrocious 720 ways!
Even if you (somehow) have every single combination figured out, there comes the need to accommodate and analyze information extrinsic to the basic task of just completing deliveries — like field agent schedules, skills, real-time traffic, and even impromptu demand. It is safe to say that the ever-growing possibilities that arose from TSP are beyond the purview of the human mind.
Where Does Data Fit in?
With the evolution of consumer needs, convenience has dethroned price when it comes to driving demand, making it all the more difficult for humans to manage the decision-making processes by themselves. For a business, that would mean settling for additional supply chain costs. Not too long ago, one could hear logistics managers talk about reducing the number of vehicles on the road for saving operating cost, but it did not work out, especially in the last mile. In such cases, the factors contributing towards operational crisis took many forms - SLAs, resource allocation, routing, traffic and weather conditions etc. In a country riddled with such uncertainties, predictions require the backing of solid data.
The Last-mile Problem
Let me give you an example. In the digital age, the last-mile delivery is about creating a personalized experience for the end-customer.
Once an order is placed, an executive is assigned to the task based on their skills, status, and schedule — all of which happens to be dynamic data. Then, they will have to provide the accurate ETA of the delivery or service to the customer with a tracking link to track the order.
Now comes the cost-saving part, where a fuel-efficient route has to be charted for the executive to reach the customer. If this sounds tedious, remember that this is for a single order assigned to a single executive. Imagine the scope and scale of a large number of orders and a fleet consisting of hundreds of such executives. It is like being stranded in a sea of data, and not knowing where or how to row.
This is where Artificial Intelligence (AI) forays into the scene, with its ability to get even the most elaborate dataset to talk sense on command. AI can leverage data platforms to process terabytes of data and create datasets to determine patterns and anomalies. These patterns are governed by predictive analysis, which can forecast the several variables mentioned in the above example.
From Robotics to Machine Learning
The AI intervention is not limited to last mile delivery. Warehouses are looking to automate and interconnect the processes within, from using Robotics for packaging and sorting to Machine Learning for optimal loading of freight containers. Location Intelligence, like geo-coding, is used to map and assess active sites that could serve as potential warehouse locations.
The B2B and B2C sector use AI-based systems extensively for optimizing routes, allocating the best-suited vehicles and resources to orders, and even the management of the field force.
Last year saw an experimental driverless truck deliver 50,000 beer cans, covering around 200 kilometres, going up to 80 km/h.
The testing served as a wake-up call for the logistics industry that employs millions of truck drivers. AI is shaping up to take over the roads as well, potentially leading to the full automation of truck fleets.
After all, we are living in a world where every activity leaves behind a digital footprint and serves as a textbook for machine learning algorithms.
It is said that logistics forms the humanities of the business world, being a perfect blend of art and science. And if you think about it, artificial intelligence aims to understand and emulate human intelligence and perceptions - striving to be as human as possible. It is about time it takes over all the human decisions involved in sending a package.