- 21st August 2018
- Posted by: Manolis
- Category: Blockchain
Artificial intelligence (AI) is playing a larger role in businesses of all types. Companies are deploying different types of AI (e.g., machine learning and deep neural networks) across a wide variety of applications, including logistics. In fact, a recent McKinsey Global Institute report indicates that AI could produce up to $2 trillion in additional value in supply chain management and manufacturing.
That huge potential for increased value was the impetus behind a recent report on AI in logistics from IBM and DHL.
The report outlines opportunities for AI in the logistics industry, enabled by adoption and advancements in the consumer space that have made voice assistant applications, for example, so common. Those advancements are already having an impact in manufacturing, where machine learning is streamlining production and maintenance processes. AI is also at the heart of how new autonomous cars can operate safely.
In logistics, AI will be part of the shift from reactive business to proactive operations. Some of this has already played out in the forecasting space, where AI and big data analytics are creating more accurate and predictive demand forecasts. The IBM/DHL report highlights four key application areas where AI, machine learning and other technologies will have the biggest impact in logistics.
Back Office AI: AI can be used to eliminate or accelerate detail-oriented and repetitive tasks. This includes “cognitive automation” to replace clerical labor; natural language processing solutions that can spot anomalies in financial transactions and automate responses; AI for automated customs declarations; and automated contract evaluation and processing.
Predictive Logistics: Analyzing a vast array of both historical data and live, unstructured data (weather reports, news reports, social media, etc.), logistics companies can more accurately predict fluctuations in demand or shipment volumes. DHL, for example, is using machine learning to predict air freight transit time delays so that they can proactively mitigate the cost and consequences of late deliveries. AI can also help evaluate traffic and weather conditions to improve route optimization and planning.
Seeing, Speaking and Thinking Logistics Assets: AI is also helping warehouse robotics systems become more intelligent—enabling them to take on picking and sorting tasks that previously had to be managed by employees. The autonomous guided vehicles (AGVs) used by Amazon and other companies in their distribution centers also rely on AI to operate safely. AI has enabled new automated visual inspection applications and vision-based inventory management solutions that can eliminate manual inventory counting.
AI-Powered Customer Experience: Retail AI solutions are also influencing how logistics companies interact with customers. For instance, DHL Parcel is using a voice-based service to track shipments using Amazon’s Alexa. Another company, package.ai, has developed its own voice agent to help coordinate last-mile delivery with customers and has reduced up to 70% of its operational costs through route optimization and successful first-time deliveries.
DHL has deployed these technologies across its operations. The company’s DHL Resilience360 cloud-based tool monitors risk in the supply chain in part from evaluating the content of millions of social media and other online resources. Using “sentiment analysis,” the tool can identify potential signs of distress that might indicate emerging risks.
The DHL Global Trade Barometer uses import/export data on commodities and 240 million variables to create several months’ worth of predictions on global trade levels.
In an article in Logistics Management, IBM explained how AI adds new layers of complexity to existing analytics efforts, because it can better evaluate the type of unstructured data (e.g., social media posts) that can affect trade or demand.
“This could be decades of pricing, decades of logistics and route management that DHL has and taking that data from physical and logistical enterprises and using AI to drive and create new values,” said Keith Dierkx, global industry leader for travel and transportation at the IBM Industry Academy. “That is really the intent here. The other thing is 80% of all the data being created today is not traditional structured data that you can find in a database. It could be computer vision, text, speech, social networks or weather data, so using the algorithms and capabilities of AI combined with enterprise data allows you to look for insights that a system continuously learns.”