Process improvement is everything in the modern warehouse, and few areas are more important than picking. Over the years, new technologies from RFID and barcode scanners to Voice-Picking headsets, have made pick and pack process improvement easier, and augmented reality is the latest technology to hit the industry.
You may recall the dismal failure of Google Glass in consumer-land, and the ridicule dispensed to those brave souls (a.k.a GoogleAsses) who drank the Koolaid and tried it out in public! Yet in the aftermath, Google and several dozen other smart glasses solutions have tapped into lucrative business niches revolving around complex human workflows in manufacturing, supply-chain and healthcare as well as other verticals. In fact, augmented reality may have a role to play in almost any profession involving people doing things with their hands while also needing continuous access to specific data feeds.
In warehouses specifically, smart glasses such as Microsoft HoloLens, free the hands of pickers to move products faster, this primary benefit is similar to Voice-Picking, but augmented reality allows Vision-Picking and much more. In addition to picking, the technology can be applied to show pickers the most efficient route to pick products for multiple tickets in one trip, to preview the object a picker is looking for, even how to pack a box in the right order. This video from Picavi GmbH showcases what’s possible.
Vision-Picking literally puts information in front of a worker’s eye without impacting their ability to pick orders. However, to fully realize the value of this new technology, Warehouse managers need to understand the limitations of their existing processes, and adopt new best practices to streamline implementation. Augmented reality is so different from previous technologies, that it is not simply a drop-in tech to replace something else. Its full benefits can only be realized in conjunction with significant process changes. But all that is for naught, without a reliable Wi-Fi infrastructure upon which these solutions depend, more so now, than ever.
Just as it is for the DHL’s of the world, the supply chain is obsessed with optimizing speed, cost and volume, and it is all about reducing steps and touches. Amazon has been at the heart of innovation and investment in supply chain systems for years, and when the e-commerce giant purchased Kiva Robots in 2012, the playing field became even more competitive.
Supply Chain logistics companies around the globe have fought to catch up to Amazon, and as a third-party marketplace, as well as a high-volume online retailer in its own right, competing with them is no mean feat. To stay in the game, continuous technological innovation is core to survival.
Amazon is building numerous state-of-the-art facilities designed to fulfill orders faster and their technological leadership is among the factors driving rivals to invest in new pick and pack technologies, and in particular in augmented reality, since it has the potential to deliver a quantum leap in efficiency. Additional driving forces for new technologies include:
Hence, IoT enabled shelving, scales, fork-lifts and people can provide the movement data to enable faster and more frequent incremental refinements to inventory management. This can ultimately save millions of dollar in saved time due to faster routes and better slotting.
Companies around the globe, such as DHL, Microsoft, Google, and Amazon, are all developing augmented reality platforms for use in supply chains and beyond. By 2017 the market was estimated to be worth more than $6 billion, the lion’s share of which exists in the enterprise sector, as shown by the following research by Xcubelab.
Field tests of Vision-Picking in logistics have been shown to reduce errors by up to 40%. Even after leaving the warehouse, automation in trucks could also reduce the time drivers spend looking for boxes and orders. Relating to logistics, up to 60% of drivers’ total time is spent searching for orders in their trucks. Reducing this cost through augmented reality could amount to 160% increase in delivery rates.
The cost of implementation varies widely from several thousand dollars to multi-million-dollar investments. The ability to achieve positive returns from such investments depends on efficiency in selection, implementation, testing, and rollout. Here are a few best practices to consider:
However for tiny IoT devices, such as sensors on a scale, Wi-Fi is not suitable because a Wi-Fi radio requires too much power. That would mean introducing new wireless infrastructure such as ZigBee, 802.11ah, BLE even LTE which adds greatly to project risk. So it’s better to favor solutions that leverage the Wi-Fi infrastructure you already have, even if the device hardware costs a little more.
Consider small devices like cell phones, typically they have only one puny antenna, maybe two versus three bigger antennas in a laptop. This affects their performance and signal sensitivity. Needless to say the smaller the device the more vulnerable it is to degraded or noisy signals.
For comfort reasons smart glasses must be made as light and as small as possible. Furthermore these are new designs, sold in relatively low volumes, with comparatively unproven firmware, versus tried and tested cellphones now in their nth generation and sold in the tens of millions. If untested, everything from sticky client issues, to roaming problems will bite you, if you don’t do your due diligence in testing and don’t have a way to detect such problems occurring later on.
Warehouses, like sports stadiums are notoriously difficult places to deploy Wi-Fi successfully. The large spaces, which may span multiple square-miles, include many reflective surfaces that can really screw-up the signal. As if a static warehouse isn’t bad enough, now combine that with moving forklifts, pallet lifters, robots and machinery. Plus shelving that one minute is empty, the next full with who-know-what material today, and maybe something completely different tomorrow. All this make for an extremely volatile RF environment - possibly more unpredictable than in any other industry.
How can you do meaningful site surveys and plan for this? Honestly, you really can’t, from one hour to the next it’s a moving target. In the past, when Wi-Fi traffic demand was relatively low volume, a volatile RF environment was not the problem it is now. The focus then, was only coverage – difficult for sure, but not too daunting with low data rates. But then came Voice-Picking technology. This led to a 10-fold increase in bandwidth requirements which demands higher data rates and higher access point density.
Now with Vision-Picking the same is about to happen, putting the onus on the Wi-Fi to deliver both facility-wide coverage and a minimum throughput SLA for any user anywhere to view a video stream in their smart glasses. That means ever higher data rates and greater access point density again. But high AP density is open spaces is incredibly difficult to control. And even harder to troubleshoot when things go wrong.
Given 802.11ac Gigabit speeds, satisfying a minimum performance SLA of say 3-4 Mbs per person in a stable enterprise setting is relatively easy. But in a warehouse, where the RF environment is continually in flux, one can easily flip from having 3Mbps one day to only 300 Kbps the next.
Of course if this happens, or worse, connectivity breaks down altogether, while a picker is fulfilling a ticket, everything slows down, and maybe that order misses a shipping window, triggering all sorts of expenses and ramifications. The secret to success is to never let that happen. So how do you do that?
This is where proactive Wi-Fi performance management from 7signal.comes into play. 7SIGNAL’s platform utilizes Wi-Fi performance sensors and agents on mobile devices to continually sample signal quality, by running a battery of performance tests. The results are all uploaded to the cloud where they are analyzed and compared against pre-defined SLAs, and then displayed on a dashboard.
The beauty of this approach, is it gives network managers complete visibility of the RF health throughout the facility from the user’s perspective, not a network-centric view. It also provides actionable alerts when performance or signal quality falls below acceptable thresholds, and it lets network managers drill down deep into the details to fully characterize the symptoms and probable cause. So they can do something about it quickly, before it gets bad enough to affect picker productivity.
Strategically distributed dedicated sensors perform more than 600 different types of tests, non-stop day and night, so they can benchmark best case and worst case conditions, and use heuristics to recognize anything out of whack. In contrast to ultra-high performance sensor access points, the Mobile Eye agent software is for mobile devices. The agent can be downloaded to smart glasses and all other Windows Mobile and Android devices used in the warehouse.
As pickers and managers move around the facility each agent periodically records actual performance its host device is experiencing in the moment, at each location a reading is taken. Aggregated data from all devices (sensor and agents) gives you the benchmarks for what performance to expect where. While individual device results make it possible to detect variance between devices, driver versions and more. This makes it a cinch to isolate problems quickly, and not be guessing as to the possible causes.
When deploying augmented reality for Vision-Picking and other applications in the supply chain it is important not to overlook the Wi-Fi upon which these solutions rely. To make sure Wi-Fi never gets in the way of maximizing the return of investment of such initiatives it is wise to implement a proactive Wi-Fi performance management system that enables network managers to detect, isolate and mitigate potential problems long before the user is affected. In the grand scheme of things, the investment required to do so, is pocket change.