“You had exceptions with 2 of your shipments (CX187J9 Princeton, CX342K3 Savannah) here is the report. All the exceptions has been resolved. Resolution Time: 3 hours. See More.” – Jonas’s mobile lock screen woke up with this notification. Sipping on his morning coffee, he sat down to explore more on what his Virtual Assistant has reported.
Back in 1956, when John McCarthy, a Stanford computer science professor coined the term “Artificial Intelligence” for the first time, little had he imagined the full scope of its unlimited possibilities. But now in 2020, having seen a whole spectrum of opportunities that the world of AI has opened us to, the story of Jonas’s virtual assistant, might not be that far away in the future. A shipment and delivery management software designed on the Artificial Intelligence spectrum that analyses, predicts, tracks and resolves shipment delays on its own without human intervention.
Picture Source: Nvidia
Through the years the evolution of Artificial Intelligence has shown the depth of possibilities that the new digital world can produce. Starting from the bigger branch of AI, the subset of Machine Learning advanced us to the world of systems that can learn, analyze and infer decisions which later evolved to the next level of neural networks through the subset of Deep Learning. Thus, began the era of a system that learns continually and adjusts the learning models automatically to respond to new and unknown situations, almost mimicking the human mind structure.
Though there is still a long way to go for reaching the perfection of human decision making, but the initial approaches and AI algorithms have shown promises in making the systems robust for the various challenges of the modern world. Given that the recent situations have turned the table to the logistics industry to innovate, ensuring high operational efficiency and creating a crisis-proof operational framework, AI might just be the answer that they need right now. The readily available network-based structure of the industry provides a good base for the AI to scale up and augment the human elements and create a highly organized global supply chain. Even Gartner had a similar prediction last year, that “by 2023, at least 50% of large global companies will be using AI, advanced analytics and IoT in supply chain operations”.
Today’s hyper-connected customer is impatient, demanding and the current markets are more uncertain than ever. There is every chance of a chaotic situation to unfurl if left unchecked. A recent study showed us that more than 80% of your customers are ready to switch brands if they are unhappy with the delivery experience. Thus experiences have become the front-runner in building loyal customers, now more than ever. And hence there is a large scope of innovations and improvements available in the shipment delivery segment where they can readily embrace Artificial Intelligence.
Let’s look into few areas where AI can paint the perfect picture for shipment delivery management.
Designing a proactive and predictive network: AI can significantly improve the delivery performances and reduce freight costs multi-fold. A learning based AI model can predict the expected delays due to operational or transit factors, related to different delivery modes and help shippers take proactive decisions on which mode of delivery channels to opt for. These decisions can turn critical cost saving measures for avoiding delays in expensive modes like air freights. Avoiding the delays can not only help shippers reduce their bills but also improve their brand loyalty with the highly satisfied customers.
Smart Route Optimization: Efficient route optimization plays one of the major roles in rationalizing the costs of last mile deliveries and driving high delivery performance rates and customer satisfaction indexes. UPS data shows that they save “10 million gallons of fuel annually by optimizing their routes”. But with that in mind, think of the possibilities that AI can bring to this required feature for every global shipper. AI algorithms can utilize its deep learning abilities and fed with the historical trip sheet data, real-time weather, traffic and environmental reports, combined with the context analysis of the social media streams; they will be able to predict quite accurately the expected delays or disruptions long before and also predict the most optimal route. The driver can be notified and they can view the modified routes on their navigation system with live maps. It can improve the overall performance of delivery runs multi-fold and create a smarter, connected and more productive delivery management framework.
Cognitive Exception Handling: Shipment delivery exceptions are unexpected and require expedited resolutions to ensure on-time delivery but data shows that on an average most shipment exceptions take more than 1-2 days to resolve. Given the complexity of the exception and the teams involved, the resolution processing time differs. But imagine an AI driven connected delivery network with augmented processes that predicts the optimal resolution in minutes and automates the execution of the resolution based on threshold parameters to replicate cognitive decisions. It might just be answer and the new future. In last mile deliveries, incorrect address exceptions is one of the common issues that drivers face. In a past Deloitte report it was found that “up to 25% of all phone numbers and email addresses stored in digital contact applications were no longer in use”. Large scale enterprises, employ large team of data experts for their customer data cleanup activities. But trained AI models can use natural language processing to performing these checks continuously on customer address information to ensure completeness, correctness, and consistency with global and regional address formats. Customs delays is another major delivery exception that occurs due to the complex human-oriented customs brokerage processes which involves lot of effort-intensive processes, cross-references and checks of documents/forms. If looked closely, this process also gives us the biggest opportunity in terms of automation and AI trained model for speeding up the process. A cognitive AI model can be trained with legislative and customs brokerage knowledge, regulatory documents and processes, along with customer and industry guides to learn how to automate customs declarations. In case of an out-of-the-box exception, they can communicate for human intervention and based on their decisions taken, the system can learn even further to develop a more robust system for expedited customs processing. Similarly the exceptions related to on-road delays or inter-modal delays can be avoided or resolved through the AI driven models as we spoke about in the earlier points.
In the current world defined by uncertainty and volatility, AI can evolve the logistics industry operating model from a reactive action based one to a more proactive and automated one powered by predictive intelligence. So that Jonas can catch up with his much needed sleep, without worrying about any shipment exception escalations that he needs to take care of immediately. While his cognitive virtual assistant can put the necessary resolution steps in motion based on its analytical reports, learnings and pre-defined threshold limits set by Jonas.
AI is all set to redefine the shipment delivery management process in the coming years. What innovations in this field are you looking forward to?