ChrisJohnson_WebA little over 6 months ago, I had the opportunity to talk about the most famous problem in Computer Science, the travelling salesman problem (TSP), and its real world corollary, the vehicle routing problem (VRP), which remains one of the most studied problems in transportation.

The issue with these problems and others like them is that they are hard – literally.  That is, they belong to a set of problems known as “NP-Hard”, and to date, no one has found a good way of finding the best possible answer to any of them in a reasonable amount of time.  Said differently, for large instances of these problems (like a TSP with thousands of cities to visit), all the computers on earth put together still wouldn’t be sufficient to find the optimal answer in millions of years.  That’s how quickly the number of potential solutions grows and how difficult they are to solve.

So, that’s the end of it, right?  The best we can ever hope to do on problems whose possible solutions grow exponentially is to get pretty good answers?

Not quite.

Let’s start with a different problem from the transportation space.  Suppose that you have 3 loads you need to tender and 3 carriers (A, B, and C) with whom you work.  Each carrier only has 1 truck available and each carrier has a different rate for each load.  Which load should you assign to which carrier?  In this simple example, there are just 6 possibilities (A, B, C / A, C, B / B, A, C …), and it’s easy to calculate the cost of each scenario and simply pick the best one.

What happens when we double the number of loads and the number of carriers to 6 of each?  720 possible scenarios is what happens.  How about increasing the number of loads, the number of carriers, and the carriers’ capacities to what you might actually encounter in the real world?  I think you know where this is headed … We encounter a mind-numbingly large number of potential solutions just like with the travelling salesman problem.  But this time, we’re substantially less screwed.  In fact, we can usually find the best possible answer in a very reasonable amount of time (i.e. typically, just a few seconds).

What gives?

The difference is that the carrier selection problem described above closely resembles what is generally known as a linear programming problem, a mathematical model in which the constraints can be represented as linear relationships.  It’s worth noting that in this context, “programming” actually has nothing to do with computers.  The phrase was coined by George Dantzig during WWII when the solution algorithm he pioneered helped plan expenditures that would reduce costs while increasing enemy losses.

Since then, the solution algorithms have continued to evolve and commercial, software-based mathematical solvers have been created to answer these problems quickly (in terms of time) and efficiently (in terms of computational capacity).

So how do you tell the difference between a crazy hard problem that you can only get good (rather  than perfect) answers to and one in which you can find the truly optimal answer in a matter of seconds?

Unfortunately, there’s no easy way to tell the difference.  Problems like network flow optimization can commonly be formulated as a linear programming problem and solved quickly.  However, that situation can change quickly with the addition of certain side constraints or other requirements.  The same goes for driver and vehicle assignment in fleets, award allocations in procurement events, and even bin and container packing problems.

Again, all is not lost.  Equipped with this information, you can ask informed questions of your supply chain technology providers and perform more rigorous comparisons of competitive offerings.  Something as simple as “Does your solution guarantee optimal answers?” is a great starting point.  If it does, the provider should be able to explain how that’s possible.  If it doesn’t, it creates an opportunity to discuss the heuristics used to find good answers in reasonable amounts of time.  In the end, this knowledge and the discussions it fosters will help you and your technology partners build better supply chains together.

Chris Johnson, VP Research & Development, oversees all software development and supply chain engineering activity at LeanLogistics. After starting at LeanLogistics as a software developer in 2004, Chris held a number of roles of increasing accountability and served as the lead technologist of several initiatives including LeanSource™, LeanDex™ and LeanFleet™, before taking his current position. With a background in computer science, mathematics, and operations research, his unique approach to software development continues to drive LeanLogistics’ technology solutions.  In addition to his role at LeanLogistics, Chris serves as an adjunct associate professor of computer science at a local university.

The Carbon Disclosure Project puts out an A List of top performers every year.  Their 2014 report lists 190 companies.  I took a look at the 21 companies listed in the Consumer Staples category, a mix of food & beverage and retail corporations.

I was curious how many of these companies were using a transportation management system.  A TMS can serve to both reduce transportation spend and lower carbon emissions. 14 of the 21 companies that made the list are definitely using TMS:

Company TMS Provider Source
Anheuser Busch InBev JDA Press Release
CVS Health Descartes Descartes Analyst Day
Danone LeanLogistics Case Study
Diageo Plc GT Nexus (Ocean) Diageo Scope 3 Logistics CO2 emissions Report
Heineken Eyefreight Press Release
J Sainsbury PLC Paragon Software Systems Case Study
Kirin Holdings Co Ltd Using TMS, solution not named Sustainability Report
L’Oreal UltraShip TMS Article
Morrison Supermarkets Paragon Software Systems Article
Pick ‘n Pay Stores Ltd One Network (used by Pick n’ Pay’s 4PL Imperial Logistics) Press Release
SABMiller JDA, Transwide JDA Press Release, Transwide Press Release
Sonae MC Using TMS, solution not named Sustainability Report
Unilever plc LeanLogistics Press Release
Wal-Mart Stores, Inc. Ortec Article

2014 Climate Performance Leadership Index:  Consumer Staples

A few warnings about this chart are in order.  First, there are companies that made the A-list that do not appear to have a TMS from an Internet search; that does not absolutely prove they don’t.  Secondly, the TMS solutions listed are not necessarily the only TMS solutions used by those companies.  I know, for example, Walmart is using other TMS solutions, they just don’t come up in the search.  It is also not uncommon for big companies to use different TMS solutions in different regions of the world.

On the face of it, it seemed crazy to me that companies committed to sustainability would not use TMS, particularly in this industry where transportation as a percentage of sales is higher than in many other industries.  By reducing empty miles, finding mode optimization opportunities, improving the transportation network, and loading trucks more fully a company not only save money, they get the added bonus of being able to report that they are committed to sustainability.

Also, documenting your carbon footprint is no trivial task.  A TMS makes it much easier to gather the base data.  Some TMS solutions, MercuryGate,  LeanLogistics, JDA, Manhattan Associates, and Descartes for example, have functionality that automates the task of calculating the greenhouse gas emissions.

TMC does as well, but they have gone further. Accurately estimating emissions from LTL is not easy.  They have created a new methodology for estimating carbon emissions of individual LTL shipments while considering the complexities of a typical LTL network.

As I looked through the sustainability reports of the “A” listers, a few explanations emerged for why some may not be using TMS.

First of all, for most companies C02 emissions from other parts of their supply chain represent a bigger source of carbon emissions and could be seen as a bigger initial target for reducing carbon.  In Heineken’s case, they achieved their 2015 targets for CO2 reductions in production and cooling, but admit they have more to do in distribution.  From a prioritization perspective, they have approached the problem intelligently; they have focused on the bigger impact areas first.

Heineken CO2 Footprint

Heineken Carbon Footprint

Secondly, a company’s ability to set long range goals depends upon where they are in their journey.  Associated British Foods made the A list and they don’t appear to have a TMS.  However, their sustainability journey is just beginning.  This company’s first sustainability report came in 2013.

Companies early in their journey are more likely to focus on what corporate responsibility professionals call “scopes 1 and 2” and defer “scope 3 until later.”  Scope 1 includes direct greenhouse gas (GHG) emissions that occur from sources that are owned or controlled by the company.  Scope 2 accounts for GHG emissions from the generation of purchased electricity consumed by the company.  Scope 3 emissions are a consequence of the extended value chain activities of the company, but occur from sources not owned or controlled by the company.  So for example, if a company outsources the transportation planning function to a 3PL, they would not control those functions and they do not own the transportation assets.  Coca-Cola HBC AG is an A lister, but they have chosen to target scope 1 and 2 GHG emissions associated with their plants.

Nevertheless, if you are a company that publishes a sustainability report, and carbon reduction is one of your corporate goals, implementing a transportation management system will usually represent an easy win.

JToday I would like to take the opportunity to briefly discuss a financial economics concept and its effects on the real economy, including the logistics market. This concept is called the J Curve. In particular, I am referring to the recent loose monetary policies of the Euro Zone and Japan. They have caused a depreciation in those currencies with respect to the dollar. Initially, the volume of US exports will remain fairly constant, but the price (in Euros or Yen) will increase. This will (or is currently) make US produced goods more costly to Europe and Japan, leading them to reduce the volume of goods they purchase from US companies. These changes, both exchange rate movements, followed by trade volumes, can have notable effects on trade and logistics. It is also makes our job as analysts forecasting market growth that much more difficult.

Now on to this week’s news:

chinaChina’s logistics sector increased its revenue in 2014 by 6.9 percent.  Meanwhile, the total value of goods transported by the sector rose 7.9 percent. The increase in logistics revenue was attributed to the e-commerce boom. China’s online sales of consumer goods increased 49.7 percent from the prior year. Now that’s a growth rate worth mentioning.

The Federal Reserve reported that US manufacturing output increased 0.1 percent in March. Business equipment was the only major market group to post a gain, and this was primarily due to transit equipment. The Bloomberg article noted that the strong US dollar is hindering growth of the manufacturing sector (see my intro paragraph above). Also worth noting, oil and gas well drilling decreased 17.7 percent last quarter. Ouch!

Cass Freight Index (both shipments and expenditures) for March increased from February, but declined from March 2014. There was mention of the West Coast ports and the effects that issue on freight volumes. It’s that same old lingering port story, so I’ll spare you the details and link to the Cass Report if you desire the specifics.

INTTRA, the online ocean shipping electronic marketplace, announced that the ocean carrier NileDutch is being added to its platform. NileDutch services shipping lanes between West Africa and Europe, South Africa, South America, and Asia. ARC’s research on the global trade management market shows that INTTRA, a consortium of the world’s largest ocean carriers, is the largest ocean booking platform in operation.

Vuzix, a provider of smart glasses that my colleague Bob Gill referenced just yesterday in his article Picking With Vision about DHL’s augmented reality picking project, announced this week that it acquired two patents that complement its existing technology. One patent covers gesture controls of 3D virtual objects, and the other relates to ambient light management in see through wearable displays. Initially I was skeptical of augmented reality glasses and their use for warehouse tasks, but it appears that their usability is improving and may become a reality in the warehouse in the near future.

An Austrailian fulfillment software vendor named Temando recently raised $50 million of investment funding. I had never heard of the vendor prior to reading the related press release. However, the vendor’s focus is interesting and timely. The software helps increase cost efficiencies of e-commerce fulfillment. This is a large part of the topic I will be discussing at HighJump’s Elevate conference next week. The research that my colleague Chris Cunnane conducted on omni-channel fulfillment noted that retailers were practicing omni-channel primarily to increase revenues, with fulfillment cost and margins playing a secondary role. This software by Temando addresses margins and cost efficiencies. Here’s a quote from the press release:

(The software) lets an online retailer offer consumers options for delivery while taking into account the cost of each fulfillment option. For example, the retailer can decide whether it’s most cost-effective to deliver from an e-commerce distribution center, a store or to have a supplier ship the item directly to the consumer… The software takes into account how fast the consumer wants a product and the reliability of the shipper.


Have a great weekend. This week’s video is Centerfield by John Fogarty



Categories This Week in News
Comments (0)

Picking with Vision

An announcement from DHL a few weeks ago heralding the success of a pilot project in Europe  involving the use of augmented reality for order picking is yet more evidence that when it comes to high technology, the once humble warehouse is not being left behind.

For three weeks, 10 workers at a DHL customer (Ricoh) facility in Bergen op Zoom in the Netherlands were equipped with smart glasses (including Google Glass and VuzixM100) that provided necessary order picking information, such as aisle number, location, and product quantity, through an application developed by Ubimax, the wearable computing solutions company,


A total 9,000 orders comprising 20,000 picks were fulfilled over that time. And the result? Faster picking leading to an overall 25 percent increase in productivity when compared to the warehouse’s usual pick practice of RF scanning. As well as enjoying their new hands-free world, staff also reported they were not encumbered by the wearables – “You barely feel it once you are wearing it.”

But just what is augmented reality (AR)? Whereas virtual reality (VR), its better known cousin, completely immerses the user in a computer-generated environment, augmented reality, as its name implies, maintains an existing physical reality but adds a digital element to create an environment that becomes a value-added mix of the real and the virtual.


Augmented reality enhances the physical environment by superimposing pick instructions in the warehouse worker’s field of view

As a relatively nascent technology, commercial applications of augmented reality are still rare. One of the more well-known is the so-called smart fitting room, which is attracting interest from a number of fashion retailers. Here, the shopper stands in front of an interactive screen and “tries on” clothing items i.e. she sees herself (physical reality) wearing a digital image of the selected item (virtual reality).

In the DHL application, the warehouse worker sees the physical reality of the aisles and racks in front of him just as he could if he were not wearing a head-mounted display, but this is augmented by a superimposed AR code in the form of a graphical work instruction, which appears after he scans the bar code at the storage location with his smart glasses. This code tells him where to go, how many items to pick and even where to place in his trolley.

With the pilot project complete, DHL is evaluating the operational suitability and economic feasibility of adopting augmented-reality vision picking. Meanwhile, its Trends Research team has already identified other logistics activities that could be enhanced by a judicious dose of AR technology.

For instance, in transportation, AR could enable rapid, hand-free checking of pickup loads by a driver using a wearables to quickly scan items for collection, followed by instructions on where best to place the loads inside the vehicle, and once on the road, dynamic traffic support based on real-time traffic data to optimize routes (the information shown on smart glasses or a windshield display). Another potential applications is warehouse planning ­­– visualizing new equipment and modifications against the backdrop of the actual facility.

It is likely to be some time before AR based picking comes to a warehouse near you or is adopted extensively for other logistics processes. But with warehouse owners constantly looking to boost productivity and reduce error rates in an environment of high-frequency pick operations and temporary staff, this is definitely a technology worth keeping an eye on (pun intended).

Incidentally, anyone with a smartphone can get a useful appreciation of augmented reality technology through free iOS and Android apps such as Augment and Junaio. So go ahead and get a glimpse of the future.

ship from store 1As I mentioned in last week’s Friday News Roundup, I spent some time in Atlanta attending the Home Delivery World, Click & Collect Show USA, and ETail Show USA joint conference. I had the pleasure of catching Tom Barone, VP of North America Operations at eBay Enterprise present on the changing nature of commerce. Tom basically outlined three key strategies for omni-channel success: ship from store, in-store pickup, and in-store associate ordering. Each of these strategies has a distinct purpose in today’s retail environment, and each one can be aimed at specific groups or types of shoppers. All three strategies have tangible benefits as well as unique challenges. Tom used his time to offer some advice on how to become omni-channel ready. The key is to start with one strategy. Tom’s advice – prioritize, and start with ship from store. According to Tom, when analyzing these strategies, ship from store falls right in the sweet spot from a “business impact” and “Ease of implementation” standpoint. Although, I will share a case study in the coming weeks where an Indian retailer opted for click and collect first, as the logistics behind delivery was too big of a challenge.

Providing an omni-channel environment for customers is no longer an option, it’s a necessity. According to data from eBay Enterprise’s customers, it is also quite lucrative. On average, retailers that are shipping from the store can achieve a 20 – 40% increase in incremental e-commerce revenue and 30%+ margin increase on markdown items. How? By leveraging in-store inventory, retailers have more items available to their e-commerce customers. This puts them in the position to sell more. And when items are reaching end of season, or simply not moving in the store, they can be sold online at full price. These are the benefits of ship from store, but the big question remains – how do you execute on the strategy? Tom outlined four keys to executing ship from store.

  1. Inventory availability
  2. Order routing capability
  3. Store operations and training
  4. Compensation and sales credit

Inventory Availability

InventoryShip from store will only work if inventory is available. Tom outlined five principles for ensuring inventory availability. First, there has to be a single view of network inventory. This is where distributed order management can help. It gives a single view of all available inventory, and allows all fulfillment channels to access the order information and pull the appropriate inventory. Second, the company must select products and categories for ship from store. The categories need to be those items that make sense to ship, from a convenience and financial standpoint. There has to be accurate store level inventory. This is generally the toughest part of fulfilling e-commerce orders from the store. Misplaced items, especially apparel, can make inventory visibility nearly impossible. If the store associates cannot find the inventory, it cannot be sold or shipped. Fourth, safety stock levels need to be set. Unless it is a dark store, the brick and mortar location is there to serve the physical customer first. Safety stock levels need to be established to make sure that stores are not shipping items that could be sold in the store. And fifth, there need to be automatic adjustments of “available to purchase” inventory levels. This means that items sold in the store, and shipped from the store, need to be applied to the on-hand available to purchase level.

Order Routing Capability

When it comes to order routing, Tom outlined three keys for intelligent routing. First is the proximity to the delivery address. There are a few different ways to fulfill from the store. The store can ship the item through a parcel carrier, or the store can deliver the item using its own employees. The proximity to the store will certainly play a role in which method is chosen. For shipped items, location will affect the delivery timeframe and expectations need to be set with the customer. Second, the retailer needs to minimize split orders. Split orders occur when a location can only fulfill one part of an order. This leads to items arriving at separate times, which some customers are not happy with. It also increases overall shipping costs for the retailer. By using the entire store and warehouse network, retailers should try to fulfill orders in their entirety. And third, retailers need to use inventory optimization and set daily store limits. Setting daily limits is important to make sure that store associates can focus on the other work that they need to get done.

Store operations and training

store opsThere is a lot that goes into preparing a store to execute ship from store. Tom outlined some of the biggest challenges and requirements to move forward with ship from store. First, there are store set-up, protocols, and procedures that need to be put in place. This includes things like what area of the store is used for ship from store orders, who is responsible for picking orders, how do the orders get packaged and shipped, among others. Having these guidelines in place is the first step. Second, retailers need to have staffing strategies. If there are specific times when ship from store activities pick up, more associates need to be on the floor. Third, the staff needs ongoing training. Training is a key component to keeping employees engaged. Making sure each associate knows the proper procedure for picking, packing, and shipping for ship from store orders is crucial for making sure it is done efficiently. Fourth, the retailer needs to have reporting capabilities. This way the company can see the actual impact of establishing ship from store processes. It also enables the manager to make more informed decisions when it comes to staffing levels, ship from store timeframes, and ship from store inventory availability.

Compensation and sales credit

Tom indicated there are two key components from a compensation and sales credit standpoint for ship from store. The first is goals and incentive plans. Laying out goals for store associates gives them something to aim for, whether that is a dollar amount or a time goal for preparing ship from store orders. Incentive plans, besides paying store associates more money, give them a sense of buy-in, or the feeling that the company is investing in them. This can be a huge bonus when it comes to fulfilling e-commerce orders. The second piece is credit for sales. The fight for attribution has been a long battle between in-store and online teams. The key is to give credit for sales to the store that is fulfilling the order.

In conclusion, Tom Baron outlined three key strategies for doing omni-channel rights: ship from store, in-store pickup, and in-store associate ordering. According to Tom, ship from store is the best place to start on the omni-channel journey. It enables retailers to increase e-commerce revenue exponentially while improving margins on items nearing markdown. To achieve ship from store success, there are four keys to keep in mind: inventory availability, order routing capability, store operations and training, and compensation and sales credit.