Augmented Reality (AR) article originally published by, and permission to publish here provided by, Robert Seward.
Augmented Reality (AR) enables transformational change!
When I started in the distribution space, I was told “…business is simple…cases in, cases out.” Reports were on green bar paper and forklifts were everywhere. As the distribution space evolved, the reports became spreadsheets and forklifts remained very important as they are coupled with the latest and greatest automation.
Supply chain improvements over the last 15 years have been incremental at best. The advances we are making in the supply chain powered by Augmented Reality (AR) are going to be wildly disruptive. The supply chain community that leans into AR and gets it right will come out the clear winners for every perspective: simplicity, safety, quality and productivity.
Starting out in an ambient distribution center
Most people know the stuff they buy at the store is not made there and would likely come from a manufacturer. How it ends up on the shelf can be a mystery. How do the products get there?
I found this out during my first job out of college. I started my industrial engineering career in a hot Southeastern ambient distribution center (DC) where innovation consisted of adding conveyor and sorters at several hundred dollars per FOOT! Yikes!
The onboarding process was faster than I could have imagined, but the operators helped me understand the day to day goals. My first boss told me, “the distribution business is simple…cases in, cases out. Corporate hired you to come down here and simplify our business…you better not make it more complicated.” He told me this during our first one-on-one meeting during my first week – what a way to start. I will go into more detail about some of his past experiences with engineers coming in to support operations at a later time. I did hone in on “…business is simple…cases in, cases out.”
During that first week I was admittedly intimidated. At the time, I graduated, drove my car with a trunk full of clothes across the US to get into a business I was classically trained for. I was enamored by the miles of conveyor belt, the scale of the operation – sorters, feeder lines and the fork lift fleet utilized on each shift. Frankly, I was blown away by how many cases were received and shipped out.
As my anxiety built, I was able to anchor on the fact that we shipped mostly cases. I focused on learning that first. The process was straight forward: trailers backed up to the dock, folks unloaded them, hauled the pallets away and then stored them. In parallel, another group was fulfilling orders placed by the store and were placing each case on a conveyor belt. The conveyor belt transported the cases to the outbound department where there were folks set up to load the cases in the trailers.
Early in the week, we did not ship out as many cases. Later in the week, the shifts were so action packed you had to be mindful of where you walked. I quickly realized that as the DC got busier with increased volume, operation meetings and updates did not seem as intense, folks were even smiling and looking forward to explaining performance numbers – volume hides inefficiencies.
Getting busier yet more efficient seemed counterintuitive at the surface but what did make sense was the busier we got, more of a percentage of cases shipped via the pallet channel. Looking at the type of loads coming in and how we received them on busy days allowed us to forecast the productivity higher.
Doing the math, handling pallets was ~4 times more productive to handle as compared to handling cases. For the engineers reading this, of course the rate depends on several factors such as: size of the case being shipped, sorter speed, spacing gap between cases, cases per pallet, pallet drive length, congestion, etc.
Learning the business
As the business demanded newer products to be sold at the store, I noticed we did not always get rid of one when we added one? That is when I learned about our breakpack channel. Breakpack was set up for fulfillment of lower demand type products. This channel allowed us to increase our SKU assortment without taking up additional fixed slots reserved for higher velocity type items.
Breakpack orderfilling was similar to case orderfilling but added several manual steps like: cutting the boxes, extracting the inner picks, rebuilding a case with inner picks and transporting the goods with pickers pulling cage carts. Note, to leverage the conveyor belt, the case had to be a specific size – individual picks were too small. With all that extra manual handling, I calculated the cost per channel for the breakpack operation.
I used a simple cost to serve calculation for the activities within the DC like: wage rate for each job task, converted pallets into cases, translated average pick count into cases, summed the cases shipped out and summed all the manual touches (labor).
As the business demand increased, the small dedicated channel that was set up for tooth paste and deodorant was ever so increasing. The breakpack channel contained the most manual steps and therefore was the costliest.
Again, for the engineers, evaluating a cost to serve is typically more comprehensive, especially at the supply chain level considering: activities within the DC, processing an order, transporting the goods, and maybe adding cost associated with capital expenditures. As I added to my flow model, the takeaway was clear: flowing cases through the breakpack channel cost more than flowing cases and flowing cases cost more to handle than moving pallets.
Another thing I had to learn to appreciate was that every report printed out on these big papers that had “green bars” on them. One report could be 40 pages! A lot of the reporting available at that time was powered by these “green bar” reports and analysis consisted of using several excel spreadsheets.
Strategic planning was a year out, so predicting manpower for the next three months was an easy task. Tactical execution was even simpler. Operationally, I would talk to my team about safety, quality and productivity. In the morning meetings with my leadership, I would translate that to “business talk” speaking in terms of accident free days, fulfillment accuracy, and cost to serve.
Distribution space 15 years later
15 years later, I found myself back in the distribution space. This time around it was in the perishable distribution center network. Why perishable? I had a boss that I worked with on an innovation team that pulled me in the perishable space because he thought it would be a fun challenge to automate.
Perishable DC’s moved cases like ambient DC’s but with a LOT more air conditioning… Learning the different perishable chambers, door rotation systems as well as the overall labor model provided me with plenty of new learnings. One constant I was able to anchor on was reporting. Reporting was essentially the same but we were able to sunset the green bar reports…
This time the fork lift fleet was divided into two groups: manned vehicles and auto-guided vehicles. Fork lifts clearly had a valuable purpose still. Forklifts allowed the operation to quickly flex our staffing up when needed. One major difference was each perishable DC seemed like 5 mini-DC’s in one building shell, especially when it came to calculating capacity as function of storage availability and flow.
Machinery was specific to each environment. The lubricants used in the forklift that operated in -20 degrees was different than the lubricants used in 55-degree chambers. One more big difference that seemed to evolve was product type. 15 years ago, I toured a perishable DC and remembered a lot of meat and potatoes. Now the demand seemed to shift more toward different yogurt varieties and coffee creamers.
With many automation advancements like automatic storage and retrieval systems (ASRS), goods-to-person shuttle systems, as well as other innovations – fork lifts still have a notable presence and reporting was still very reactive versus proactive.
Analytics and simulation helped pressure test material handling solutions before we spent several million dollars to procure the solution, but insights were something you heard about to better understand consumer behavior.
Business challenges for me grew from the normal set I was accustomed to now included new attributes: capacity challenges, job force availability, system flexibility, system scalability, commitments to reduce the replenishment cycle time, increase asset utilization, reduce inventory levels at each retail unit and increase assortment. The challenges were very complex and were FUN to attack. In a future article, I will walk through the different solution suites created.
Many of the fun memories of implementing automation in a perishable environment presented challenges and frankly were not fun during the install time…lots of hard lessons learned! We course corrected some logic and changed the design on some of the braided wires and delivered a great product. Even with all the fancy automation to reduce the cost to serve and better utilize space, we still could not fully automate. The fact that we still used fork lifts and continued to collect enormous amounts of data, what would it take to create a step change improvement?
Augmented Reality (AR)
I believe computer vision and Augmented Reality (AR) coupled with Artificial Intelligence (AI) will create that step change improvement. Admittedly, I am not an expert in the AR or AI space, but when I was able to find colleagues that were experts in the sensory space and we partnered on a specific supply chain pain point, me being the business, I wanted to create an intelligent supply chain (end to end).
I presented my supply chain opportunity to them and I knew I got them interested when they asked if we could schedule a white board session later that day. The collaboration of business expertise and technical innovation was the perfect recipe! I represented the business with my supply chain experience and they brought their enthusiasm and technical expertise.
Together my business case went from something that I planned to research and attack over a 2-3 year horizon to a functional AR prototype within a couple quick months. I outlined my 5-D framework (discovery à define à design à develop à deliver) and we started later that week.
Of course, I was excited about this solution being brought to life. Anyone in the business that secures technical resources is always pumped up – get as much done before they get pulled from the project to focus onto a different initiative. I will discuss the rapid prototyping process and how cheap the solution ended up being in another article. I want to cover some key learnings that I believe will transform the way of working in the supply chain domain.
Regarding the Augmented Reality (AR) product, we built algorithms to create calculations to redesign direct labor out of the system plus built a backend process to remove non-value added time associated with set-up and wayfinding. My newly formed team proved to be very disruptive. We went from people telling us the AR technology was not ready yet to creating a tangible roadmap. Honestly, I was blown away how simple and intelligent we made the solution powered by AR.
Implementing the solution and the change management associated (learning curve) took minutes versus weeks. The operators quickly adapted. We created a step change improvement regarding operational simplicity, quality, and productivity.
One of the biggest “aha” moments came from our very first cycle run. As we walked through the backend processes, to perform some technical quality checks, we noticed the data capture from this single cycle was incredible. After my team created what we thought was a relatively “perfect” operational environment (from a control standpoint), the AR solution picked up on peripheral errors as well as generated a very detailed time study.
For the engineers, no more motion elements software and no more stop watches. With several cycles, we tied the AR ability to capture peripheral errors with the inventory system, and attributed specific elements into value added and non-value added categories. From there, every cycle run was dynamically calculated (in real time), as well as updated the system in the background without operator intervention. To think we almost skipped over this data file, as we were more focused on the hardware and software integration, was wild.
Takeaway – AR is the Future of Supply Chain
When I started in the distribution space, I was told “…business is simple…cases in, cases out.” I clearly understand what my first boss said, and he was right. The business should be made as simple as possible.
Streamlining everyday tasks, performing wildly complex computations, and having a personal assistant to talk you through exceptions will be a staple. The future of the supply chain is going to be fun ride!
It may be difficult to imagine how something as simple as intelligent glasses will reshape new and existing supply chain networks, but I know it is possible. I wonder who will lead the charge in the supply chain space…