This post looks in depth at how different approaches to implementing voice technologies in a DC determine the magnitude of order picking productivity gains any facility may achieve. Use these guidelines to estimate the potential results in your DC.
What Are The Three Strategies?
It would be nice if we could make precise, blanket assertions about the order picking productivity benefits of voice. But it just isn’t possible to say that DCs moving from paper to voice for case picking will always see 10 percent productivity gains.
In reality, a facility’s current technology and pick process are only a part of the equation. The biggest variable is how you approach your voice implementation: as a voice-enablement technology project; a workflow improvement project; or as a process optimization project.
Companies that treat voice as a technology upgrade alone tend to see 5-10 percent productivity gains, while companies that treat voice as part of a broader warehouse operations optimization initiative see gains ranging from 25-35 percent in case picking productivity, and 45 percent or more in piece picking.
1) Voice-Enable an Existing Process
Voice-enablement substitutes voice direction and speech recognition for visual displays, scans and key entry in a WMS-directed RF process. Productivity gains result from eliminating the time spent stopping to read device screens, pausing to handle a scanner, or slowing down to key-enter data or press function keys.
It is also common today to combine voice, scanning and displays in a multi-modal process which allows DCs to use the best tool at every step in a process – for example, allowing order pickers to scan barcodes when that is more efficient or accurate than voice or key entry.
In general, voice-enablement focuses on improving accuracy and efficiency at the pick face, with little change in travel time or other aspects of the workflow. Typical picking productivity gains are in the single digits.
By The Numbers: Voice-enabling an RF application will save 1-3 seconds per pick versus a typical RF workflow. In a simple picking scenario with a current pick rate of 100 lines/hour a two-second time saving per pick would improve the pick rate to 105.8/hr., a 5.8 percent productivity gain.
2) Workflow Improvements
To eke out additional order picking productivity, DCs should look for ways to change and optimize their workflow to eliminate wasted time and unproductive steps. In addition to substituting voice for RF (or paper) and leaving the process unchanged, you can condense or combine process steps or change other aspects of the process flow, streamline exception processes (which are extraordinarily time-consuming), and reduce the time pickers spend doing tasks ancillary to the main workflow.
For example, you can shave seconds from every pick by combining multiple voice prompts (“go to location ABC,” and “pick X”) and user confirmations into a single step. Next, to reduce travel time between assignments, you can allow pickers to start their next assignment where their previous assignment ends, rather than walking back to a starting bench.
Similarly, most DCs have opportunities to streamline pre-pick set-up time, post-pick staging tasks, and to reduce time spent handling exceptions (stock outs, wrong item in slot, etc.). Building in these workflow changes can eliminate wasted minutes from every work assignment.
It’s important to note that companies using voice as part of a work execution system can implement these types of workflow improvements without making any changes to their warehouse management system (WMS), slotting, or other material handling systems.
By The Numbers: Dialogue and task optimizations compound the benefits of eyes and hands free activities, shaving extra seconds from every pick, with additional minutes saved in exception handling, reduced travel, and in pre- or post-assignment steps. With the same pre-voice pick rate of 100 lines/hour, a four-second time saving per pick would net a productivity boost of 12.5 percent. A three-minute savings in set up or post-pick time (assuming one assignment per hour) could add an additional 5 percent productivity boost beyond that.
3) Process Optimization
To get really dramatic levels of efficiency, DCs can reengineer their processes as they implement a voice-directed warehouse optimization solution. This could take the form of introducing dual-pallet picking instead of picking a single pallet at a time, or, if you are already picking multiple orders in a batch, batching work assignments differently to optimize pick density and reduce travel, or moving to a zone picking process.
Other process optimization examples include task interleaving, two-stage picking for slow-moving items, or developing a bucket-brigade pick-and-pass process to evenly distribute work in a pick module.
DCs that need to capture and track item level data prior to shipping (lot, serial, etc.) may want to add a data capture step (either by voice or scanning barcodes) within their picking in order to eliminate or streamline downstream data capture processes. Like process optimization, these transformative changes can typically be made without changing back-end systems using a work execution solution.
By The Numbers: Process optimization can result in high double-digit productivity gains, depending on a number of factors. In our 100 pick per hour scenario, moving from single-order picking to picking two orders at a time would effectively double the pick density (i.e., the number of picks per aisle, or area), cutting travel time per pick anywhere from one-quarter to one-half. (Even if you were to cut travel distance in half, net travel time would be reduced by less than that as the order pickers would be making more frequent stops and starts along the way.) Doubling the pick density and optimizing the pick workflow should generate a minimum 10 second time savings per pick for a net 38 percent order picking productivity gain, not to mention the potential time savings from streamlining other tasks. Frankly, this is a fairly conservative estimate.
Calculating the Gains Given Two Different Scenarios – Case Picking and Piece Picking
The following two examples illustrate how the different approaches to using voice can translate into distinctly different levels of benefit in simplified case- and piece-pick scenarios.
1) Case Pick To Pallet
The table below illustrates the impact of the different implementation approaches in a hypothetical DC with an average RF pick rate of 100 cases per hour. To keep things simple, we are assuming there is a single case picked per line and that workers are picking a single order at a time to a single pallet.
In this scenario, voice-enabling an RF application would save 1-2 seconds per pick versus a typical RF workflow by eliminating scanning and screens (column 2). Dialogue and workflow improvements (column 3) compound the benefits of eyes and hands free activities, shaving additional seconds from every pick. Additional minutes per day can be saved in exception handling, reduced travel, and in pre- or post-task steps. The table, below, only includes in-task workflow improvements.
In the Process Optimization column (column 4), we are assuming the DC is moving from picking cases to a single pallet to picking to two pallets at a time. This effectively doubles the pick density (i.e., the number of picks per picking assignment), cutting travel time per pick anywhere from one-quarter to one-half. Doubling the pick density and optimizing the pick workflow would conservatively save eight seconds per pick for a net 29 percent productivity gain (not to mention the potential time savings from streamlining other tasks).
2) Piece Pick to Cart
Our second scenario is based on a piece-picking process in which multiple customer orders are picked as a batch to a cart. In this example, we assume an RF pick rate of 200 lines per hour.
Similar to the previous example, voice-enabling an RF application would save 1-2 seconds per pick, but given the higher number of picks per hour (and less travel per pick), the magnitude of a one-second per pick savings is far greater than in our case-pick example. Likewise, there are larger opportunities for workflow improvements in a batch picking scenario, such as combining picks for multiple orders in a single transaction (pick and deal).
Adding to the productivity gains, it is possible to improve pick density and reduce travel without changing the number of orders in a batch using mobile work execution. For example, by batching orders that include items in the same aisles and bays can eliminate so-called “empty” travel. Likewise, applying pick-path optimization logic can reduce travel 20 percent or more overall, driving the total productivity savings to greater than 50 percent.
How to Determine Which Strategy Is Right for You
As illustrated in these two simplified examples, it is fairly easy to identify and quantify the time saving from voice-enabling screens and scans in an existing RF process. It is also straightforward to project the productivity benefits of streamlining or improving an existing workflow. It is a bit more challenging to project the potential gains from process optimization as that requires a clear picture of what an optimal process will look like in your DC.
A good way to get started in developing a vision for that end-state, and for projecting the potential efficiency benefits of new picking methods, is to conduct an operations assessment. Similar to a lean assessment as part of a six-sigma process, you will need to document how and why you are doing things today to identify specific time-saving opportunities in your existing process. Beyond the possibilities for process optimization, the assessment exercise will typically suggest new ways to better achieve your business and operational goals – it forces you to think beyond “this is how we do things” to “this is why we do things.”
Assessing Your Current Operation
To help DCs conceptualize their own optimal process – and to begin to estimate your end-state productivity gains – Lucas offers an operations assessment service. The purpose of the assessment is to identify specific process improvement opportunities in your facility. Many of the process improvement ideas that come out of the assessment will involve the application of new technologies (voice, scanning and device displays, etc.) in a new user workflow, along with other optimization technologies (such as batch algorithms and pick-path optimization engines). In some cases, process improvements can be implemented with no change in picking technology, but voice and other new technology (RFID, pick to light, etc.) may compound the benefits of any process changes.
This exercise takes some time, but it is time well spent. The assessment allows you to consider how new optimization technologies and workflow solutions can impact productivity beyond the efficiency gains of voice at the pick face. Taking this approach can translate into hundreds of thousands of dollars in annual labor cost savings.
