Monday, March 02, 2015

Tutorial: Using Labor Standards in the new Warehouse Management solution

This post consists of two parts: An introduction to labor standards from a business standpoint, written up for me by Steve Mulaik from the Progress group, and then a short example of how all of this looks in AX today.

Introduction

One of the less well known but very distinguishing features of the new Warehouse management solution in Dynamics AX 6.3 is its support for multi-variable, engineered labor standards.  Labor standards are very important in warehouses that have lots of people.  A common misconception is that they are used to increase the pace of the workforce.  In reality, pace rarely increases much if at all.  Two other positive effects result from the proper use of labor standards.

First, the utilization of the workforce improves, i.e. the time when workers are working.  With labor standards a warehouse manager can better estimate how many people are really needed to do a job and he can schedule accordingly.  A surprising outcome from standards projects is that they reveal how much supervisors “hedge” labor to cover the unknown.  Supervisors are much more willing to move or trade resources from their area to other areas after labor standards are implemented; they can see how much their department’s performance suffers if they do not.
AX can provide feedback on historical performance for a given warehouse task and that information can be used to develop a good planning number for how much a person can get done in a given area in a given segment of time.  No one can remain busy and perform well in a warehouse or warehouse area that has 30 people that really needs 25.


Second, labor standards can drive improvements in productivity through better methods.  A standard should be set around the best way to do a job.  A lot of companies have no idea what the best way is to do a job until they do a labor standards project.  Once identified, companies teach everyone the best methods and measure each person’s productivity.  Associates that are not meeting performance often times are not following the proper procedure.

What is a labor standard?

At this point, you might be wondering “What exactly is a labor standard?”  A labor standard is a formula used to calculate the estimated time it should take to perform a given class of work.  Labor standards are derived from timestudies of associates actually performing the work.  A timestudy is used to calculate statistically valid estimates for each element of a given class of work.  An “element” is a step a worker performs when doing a job.  For example, in picking there is an element for “Scan location” to confirm that the picker picked from the right location. 

You can see an example of how a timestudy looks and is then used here.

Once the timestudy is complete, as a result, we are getting out estimated times for executing a task, for example, 13.29 seconds for every line the user picks and 3.57 seconds for every unit he picks (See above timestudy example for more details).

It is worth noting that in many other systems this formula can only have one variable, e.g.
Estimated Picking Time = 18.26 seconds * # of lines picked
but in AX we can support two variables per standard (One for the Pick and one for the Put).  This means standards implemented in AX in most instances will be more accurate than in these other systems.

Labor Standards in Dynamics AX

So, let's use the results of the timestudy we got above (or one like it), and configure labor standards in AX, and then execute some work and see how the actual time is calculated accordingly

First thing we need to do is choose, which warehouses we want to enable for these calculation (note, that since they are done for every work line created/executed, you should only enable it on those warehouses where you actually need the data).

Labor standards configuration
Allow labor standards on Warehouse

The story behind the example

Our company sells fancy speakers. They are fancy, because they are relatively small, and are usually sold in cases, where one case holds 10 speakers. The orders are typically for 50-60 speakers (or 5-6 cases), because that is how many are installed in the smart houses they are made for. Here's how the product definition looks in AX:

Released product details
Released product details - FancySpeaker

The warehouse workers do case picking, going out to the picking location for the fancy speakers, picking up the 5 cases, stowing them on the pallet on their forklift, and then delivering the pallet for that order to the outbound dock, where it is loaded onto the truck. 
The work template supporting such an operation is shown below:

Work templates
Work template setup

The warehouse worker (John) based on our timestudy needs 36 seconds in total to drive to a selected location (about 20 sec) and then, after picking, to the dock to deliver the goods (about 16 sec). At the picking location, he needs about 9 seconds per case to pick it up from the location and put it on the pallet on the forklift. 

Sammy, the shipping clerk, who does the loading in our company, needs about 7.5 minutes per pallet (with 5 CS on it) that he needs to load on the truck.That is based on the volume of the pallet stacked with cases.

Labor standards configuration

We configure these times in AX using the Labor standards form, available under Setup in Warehouse management module.

Labor standards
Labor standards configuration

This is pretty flexible, allowing for separate configurations based on the Work trans type such as Sales and Purchase orders, Replenishment, etc., and Work classes, meaning that each unique warehouse operation can be timed separately. On top of that you can set up filters per Item/Location, to which the standards should apply. 

All the configuration values are specified in a uniform fashion, as X units per hour. This could be 
  • X transactions per hour (which is equivalent to X locations visited, as each work line corresponds to a specific location), where I specified John needs 36 seconds per location => 3600 sec (1 hour) / 36 seconds = 100;
  • it could also be in specific units, like in my example with 9 seconds per case (3600 sec / 9 CS = 400 CS per hour);
  • it could be based on weight/volume, like in my example, where Sammy loads pallets with approx volume of 0.5 m3 (based on product volume information and typical order quantity) in 5 min each => 6 m3 per hour (see note below);
  • it could also be work units, which could be applicable to my example as well, because we know the order is usually just the 1 pallet, so they would have 1 work per order, and thus we could base the calculation on the number of work orders processed per hour. 
The Work types we can select are Pick, Put, or Pick and put, which is useful for companies who don't want to spend a lot of time measuring each part of the picking process, but want to estimate the time for the entire operation at once. 
Note: I have used the Pick&Put for my Load work class, and have doubled the estimated work per hour (12 m3 instead of 6 m3 I calculated above). That is due to the fact that AX will stamp the individual work lines with the time, virtually doubling the estimated time for the operation.

Walkthrough for estimating the work for a sales order

Our company received an order for 5 cases of FancySpeakers from one of our frequent customers, Forest Wholesales, and don't ask me why they need speakers :).

Sales order with one line
Sales order for FancySpeaker

We have sufficient inventory in our warehouse, so we have accepted the order and are starting the picking process

Inventory on hand
Available on-hand

I will skip the part where we release the sales order to warehouse, add the created shipment to a wave, process and release the wave, thus creating the necessary picking work for the warehouse crew. Here's the work that will be created as a result of these steps:

Work header and lines
Work for FancyProduct order

As you can see on the screenshot above, the Estimated time for each work line has now been calculated, as well as a summed up total on the work header.
Note: This field, along with Actual time, Work in process and Closed work, are available out of the box in the product, but are not by default shown on the Work details screen. I have added them through Personalization on that form.
If we look at the calculations, you can see we are only picking from 1 location, and, as calculated above, it should take John 36 seconds to get there and then to the dock, 45 seconds to actually pick the 5 cases (9 sec per case), and then it should take Sammy 5 minutes (300 sec) to load the 5 cases on the pallet into the truck.

Walkthrough for capturing the actual picking time

In order to execute the work we will use the Warehouse mobile device portal shipped with Dynamics AX 2012 R3. For that, we need all the relevant work user, mobile device menu and menu item setup, etc. We will need to separate menu items, one for the picking operations done by John and one Sammy doing the Loading, which is also clear from my use of 2 separate work classes as part of the work template. Below are the mobile device screens John and Sammy go through. I have omitted the way they got to these. In my system I used user directed and user grouped execution modes, but in your system this could be completely different. If you are not familiar with this configuration, please refer to the following article on TechNet.

Warehouse mobile device portal Sales Pick
Mobile device screens for John
Warehouse mobile device portal Load
Mobile device screens for Sammy
I wanted to show these screens, as it is important in terms of understanding, how the Actual Time when executing work is calculated. 

When the work user sees the first of the two screens, that means that the selected work line with Work type = Pick is now in progress, its state is updated accordingly, along with the timestamp of when work execution started. Confirming the pick by clicking OK (or simply scanning in the required data with an auto-submit) marks the approx. time when the corresponding work line is finished (or, in terms of AX, closed). Same for the second screens in each image, with the only difference being that it is now a Put line.

Depending on the way your warehouse workers go through these screens, you might need to adjust your labor standards configuration to improve accuracy of measurements. 

OK, so now the work has been completed and the speakers have been loaded on the truck. Let's see what the work looks like:

Work header and lines
Work completed for FancySpeaker order

As you can see above, the Work in process and Closed work now contain the time stamps for when the work was started and when it was completed, as well as the calculated value for Actual Time for each of the work lines. The work header is also updated correspondingly. 

Reporting on productivity

Now that we have the standards setup in AX, it will calculate estimated times for each work line for each work class with a standard. Now you will want to start reporting on work user productivity. To do that, we need to introduce two more topics: employee efficiency and employee utilization.  Employee utilization measures the time that an associate was working. We can’t really see this in AX but we can estimate it. It probably is equal to the time they were on the clock minus two breaks of 15 minutes each and some time at the start and end of their shift.

Time on the Clock480minutes
Start of Shift-8
Break 1-15
Break 2-15
Personal Time-12
End of Shift-8
Off Standard0
422minutes

This would suggest that most employees are about 87% utilized on most days if a warehouse is well run. At some sites the employees are given cards to fill out when they don’t have anything to do or have to do something that doesn't have a standard and supervisors will subtract this “off standard” time from the available minutes to work when calculating an associate’s performance.

To report on associate productivity we can print a simple report that shows the total actual (whether that is captured by the system or taken based on worker utilization) and estimated minutes associated with each work user (for example, on a given day). We can then export this to Excel, and divide the Estimated time * 100 / Actual time, getting the worker efficiency in %.

Reports available out of the box in Dynamics AX

In Dynamics AX 2012 R3 there are 2 simple reports that can be printed, that show two different slices of the captured execution time data.

SSRS Report: Labor by work transaction type
Labor by work transaction type
Labor by work transaction type could be interesting for warehouse managers when investigating inefficiencies in certain warehouse processes. They could for example see that as of this week, it takes much longer to complete picks for transfer orders and dig deeper to figure out the problem with the process.

SSRS Report: Labor by user
Labor by user
Labor by user can help us measure worker productivity, as described above. As we can see from the above report printout (of course, it is only for 1 work order, so is not really sufficient for such measurements), John needed a bit of extra time to perform his picks, which earns him an efficiency score of 1.35min/1.42min = 95%. Sammy was on the mark, getting 5.00min/4.98min = 100.4%.

Assuming that the defined standards are good, the efficiency should be around 90% to 100%. Sometimes some associates can earn over a 100%, but scores over 120% should be scrutinized because usually it is impossible for a human to outperform a well-designed standard by more than 120 to 130%. More likely if the number is low like 70% or even 50%, it usually points to a problem in employee utilization. The associates are not working all the time or there are too many people to keep busy all the time. Labor standards thus lead to changes in supervisor behavior as much as they do associate behavior. Supervisors have to run a much tighter ship under labor standards.

Conclusion

Labor standards are one of the best ways that AX can really improve a warehouse. It is in every warehouse manager’s interest to investigate using them. Even if your workforce will not accept being measured, you still could get great value from standards. While a popular opinion is that labor standards drive workers to “sweat more” the reality is the improvements in productivity rarely come about from driving up worker pace. The improvements which can be 20 to 30% or more usually come about from teaching supervisors how many people are needed to get a certain amount of work complete and in teaching associates the best way to do a job.

Try them out!