Mobile Phones/Tablets are getting used in Warehouses/Distribution Centres to scan bar-codes and show order details and perform put away, picking and packing processes. Warehouses/Distribution Centres are using IoT technology to monitor temperature inside the warehouses and also to detect presence of allergens.
Warehouses are also taking benefit of smart lights to use energy efficiently. IoT is getting vastly used in Fleet Management solutions, where drivers are using mobiles/Geo devices to effectively manage route, control temperature of the items in transit, avoid traffics and perform last mile delivery. Yard Management solutions are using IoT to communication with drivers seamlessly to optimize the positioning, loading, unloading and dispatch of Trucks. IoT enabled Asset Management functions are getting used across the Supply Chain Industries to track products, manufacturing tools, fleets and even workers and drivers. The suppliers and manufacturers are using IoT applications to hire logistics providers and transporters and similarly the logistics providers and transporters are updating the suppliers and manufacturers with appointments for pickup and accurate status of the shipment orders. Port operators or Dockworkers are using Mobiles/Tablets to track unloading and loading of cargo information in order to communicate with shipping lines/agents and logistics providers with the shipment details.
As per the reference of Ryan Schreiber they describes as follow. This is the $600 billion question, isn’t it? My background is in the domestic transportation realm, so my answer will focus mainly there. Like I have said in other answers to these topics, “logistics and supply chain” is a really broad topic. At a granular level each piece of the supply chain has a different challenge. The FTL market domestically is 600B, and its incredibly fragmented. That’s why it’s the $600B question.Framing the problem: When I was first learning about this industry, I asked the RFP Pricing Director at Echo what I thought was a relatively simple question: “If you had to pick one place in the US to set up a warehouse to which and from which to do all your shipping, where is the optimal location?” He couldn’t give me an answer. He said that he’d want to do all his outbound shipping from NJ and all his inbound shipping to Chicago. For what it’s worth, I totally agree with him within the parameters of the question I posed.
What I realize looking back on that question is that it is much broader than I was thinking. I was thinking “hypothetically, if I routed all the transportation in the US”. In that case, the answer is probably Plato, Mo. but I don’t route all the freight, and neither does Echo (or anyone else). Shippers answer this question every day in making their supply chain decisions. What makes sense today for them is a moving target based on volume, on orders, on manufacturing (if they do that) or ports of entry (if they do that), on access to trucking capacity (if they need it), on things of which neither of us are probably aware, etc.
That’s why On Demand solutions that solve for part of that calculus are starting to gain traction in the near-term (be that something like FLEXE for warehouse space, or shippers moving more toward Spot Market procurement). In the intermediate-to-long-term shippers will have to be nimble because there are two things moving independent of each other: the shippers need for supply chain solutions, and the supply chain market-place reacting to overarching trends. I had this conversation recently customers buying transportation more on the spot market. The opposing party was making the argument to me that it is a trend which will continue to a limit approaching full buying on the spot market. I argued that it’s a cycle likely to swing back toward contract. Customers are moving toward spot buying now because they screwed themselves in contracts in 2015 and 2016 when rates were declining. As the buy-rate increases, the customers will want to lock in rates which may be lower.
Predictive Supply Chain, Deep Learning, AI: What we’ve seen over the last 20 years is the development of Big Data (there’s a reason it’s a buzz word). Companies been capturing this data, but haven’t been sure how to use it. For shippers, their supply chain is still mostly a blackbox to varying degrees (just google the phrase). Every company, even Fortune 100’s with whom I’ve worked, have visibility breakdowns in their supply chain. At lot of times the black box prevents them from answering the most essential question to their business – “How do I get products to my customers more quickly for less money per unit”. Whether that’s fewer truck delays, better inventory distribution, etc., it rings true that the supply chain is a drain on the balance sheet of shippers. We’ve used some of the data to do things like “route optimization” to take (what I would describe as) a high level look at how companies are making routing decisions to help answer questions related consolidation, or routes, or the like.
In the last few years companies have gone a little further by working IoT principles into their products to take into account traffic, weather, real-time tracking, etc. into routing algorithms. Neither of these go as far as we’ll see in the coming years with deep learning. When it comes to predictive supply chain and Big Data, the industry is going to see a move toward less “guess work” in making projections on what to move, when to move it, and how to move it. Deep Learning Algorithms are going to help shippers better anticipate customer demand (Amazon is already working on this). This will mean a number of things. The first, and most obvious, is a more optimized inventory management. The second, is a better analysis of the total distribution cost per unit when accounting for where inventory will be and how it gets there (essentially, my question to the Echo guy above).
The last will be better and more flexible transportation procurement opportunities. The last one is the one that means the most to me, of course, as a transportation provider. Being able to help shippers anticipate their supply chain and provide options is best for both shippers and transportation providers. Right now, data is good at telling us where product needs to be and pretty decent at telling us ho to get it there. It’s okay with telling us how many things we might need to get there based on how many things we got there last year. What it isn’t good at telling us is when and where we will be sending those things, and how many more of those things we’ll be sending, based on external factors like market demand etc. Because currently the market is not good at these things, it leads to a number of inefficiencies, such as a build up in inventory in a location or JIT manufacturing to minimize waste.
The goal is to minimize the overall strain on a customer’s supply chain. From a purely procurement perspective, flexibility allows customers to benefit from modal-optimization and from price-optimization. The industry now is and incredibly inefficient system of matching supply to demand. Transportation providers are working with the same variables shippers are in servicing them: two markets moving independently of each other. That’s where AI, NLP, and Machine Learning are going to be the difference makers. Prospectively being able to anticipate both when and where product to arrive before a shipper’s customer asks for it and helping shippers make the decision on when waiting might be better, or moving a shipment up might be transformative to their overall supply chain benefit. Think about something like purchasing an airline flight. When a traveler is price-sensitive, but vendor (airline) and date insensitive, s/he has the maximum amount of options.
Think about Orbitz +/- 3 day fare option. It allowed the traveler to use all available information to maximize cost structure efficiency. That’s the next step for Predictive Supply Chain. Using Deep Learning on shipper data, on carrier data, on the IoT networks and providing shippers with the best choices that meet their criteria. Said another way, shippers now make decisions based on when things need to arrive and generally work backwards to when they need to ship to decide when to make (or get) whatever needs to be shipped. Optimally they can make decisions accounting for all three. In the end, change has to be coming to domestic transportation. The industry is too inefficient and too expensive. Deep Learning, AI and a Predictive Supply Chain is the way of the future.