Tugboat Software Blog

January 26, 2012

One of our General Mills customers recently re-bid 96 jobs. This was due to installation of some new equipment. When they’ve had to do this in the past, with a workforce of over 400, this was an administrative nightmare: the workers filling out individual bid sheets, sorting through the bids, ranking them in seniority order, awarding new bids. Then due to bumping, they’d have to do it over again figuring jobs that needed to be re-bid.

Tugboat’s recently launched tools have automated this entire process. A staffing level planning tool enables management to plan the number of needed jobs. It automatically posts the new jobs to a job bidding window which is available to all employees via Tugboat’s self-service portal. This allows employees to do the rest by themselves. The bids stay open for two weeks so they can also cancel and prioritize their bids. The bids are processed automatically when the bidding closes. All this with no paper!

Had they continued doing this by hand, explains the General Mills scheduler, it would have taken over 2 month.  “It made our life a lot better. Tugboat got us through the bidding process more accurately. A planner’s dreams come true”.

Does this situation occur with your workforce? When the workforce is increased or decreased in size? Or for simple open jobs when someone leaves the workforce?

Amongst pressures faced by manufacturing operations today, matching the size of one’s workforce to their projected production requirements is a growing challenge. Tugboat’s new job bidding tools include a labor planning interface that manages the desired number of jobs for each department and shift. Users can develop a list of minimum jobs required during both slow and peak times to assist in determining their desired staffing plan. Once these levels are set, available job assignments are posted for job bidding to existing employees on the employee Kiosk. Tugboat’s workforce scheduling solution will also automatically awards the new job assignments according to HR policies.

What makes this challenging for larger operations is that typically scores of employees, or more, are affected. They refer to this as ‘furlough and recall’. If people are furloughed, job assignments for the remaining workforce have to be reset. However, any job assignment changes must conform to existing union or HR-policy rules, so. There’s a ripple effect on other workers as senior workers bid for jobs now that the workforce has been reduced. Similarly, during a re-hire multiple new jobs are opened up and again current senior workers are entitled to bid on these new jobs. This again has a ripple effect.

Tugboat’s tools enable employees to bid on job assignments during a downsizing or bid on new positions when production is again increased. Tightly integrated with Tugboat’s labor scheduling solution, these tools considerably reduce the workload for HR and front-line managers by placing responsibility for job bidding in the hands of each worker. Tugboat’s proprietary optimization engine will automatically assign new job positions according to all of the Customer’s site-specific labor scheduling rules while conforming to their job bidding policies and planning goals.

We’ve also developed this new software so that it is tightly integrated with Tugboat’s skills management solution. Workers assigned to new jobs, are automatically signed up for any needed on-the-job training.

Even if an application appears to really meet one’s needs, there is always the challenge of implementation. A financial manager for Coca-Cola recently asked, “This application looks like it will do the job but answer this: how much trouble will we uncover when it comes to implementation?” He was specifically referring to the resistance that would likely occur from their schedulers and a unionized workforce.

When it comes to those that actually do the scheduling, most would rather do things the way they’ve been doing them – even if there is a little inefficiency and maybe hidden costs due to some redundancies. When it comes to the workforce, most will naturally want to avoid changes in how jobs are assigned. However assigning them more efficiently takes some analysis of the rules and policies that control how the workforce is currently scheduled.

So how do we implement a fully automated application like SOS without running into resistance from the schedulers and the workforce? Along with getting the data for jobs and people in place, a project usually starts with focus on the existing scheduling rules. However, there’s a lot more to automation than getting all the rules and policies analyzed and here’s a way forward.

Right from the start, even from the Internet at home, Tugboat’s solution enables workers to make requests for:

•    Overtime

•    Days off

•    Vacations

•    Training

•    Open jobs

This interface also makes reports easily available to the workers for:

•    Absences

•    Requests

•    Available OT

•    Scheduled jobs

This is a two-stage approach and it takes the pressure for change off of the schedulers. Automating requests and reports means, right from the start, less paper and less discussion with each and every worker. And the workforce? Automating requests and reports will make life on-the-job a little easier without raising questions about how jobs are assigned. It’s called empowerment. For management, the cost of getting to this level of implementation is fairly low. Having data for your existing workers and jobs uploaded into the application is all that’s required.

To summarize: save the challenge of analyzing and documenting all your rules and policies until a later stage. Get the easy stuff in and running for an early payback and buy-in from the workforce

In large manufacturing facilities, it’s not hard to find managers who’ll acknowledge, on any given day, there may be extra workers on the floor. Sometimes referred to as redundant manning this sounds like an easy target when looking for labor waste. But is it? Yes and no. Here’s the difference.

Under certain circumstances having an extra body on a line is a good thing. Any machine can get a little quirky and require an extra pair of hands and eyes just to get the product down the line. There are appropriate times and places when extra manning makes sense, so. Any automated labor scheduling solution must enable an override for human judgement. However when this becomes an habitual alternative to your standard manning, something isn’t right.

If say the labor standard for an operation is 10 workers, yet some supervisors are only comfortable with 11, then management has a couple of choices. Pay for the 10% premium above the standard or adjust the standard upward. In other words, if this needs to happen every day or every shift, then change the labor standard to 11.

A solution that allows the scheduler to make special assignments, assignments that typically fall outside of the basic manning for building the schedule for a particular product, will highlight these non-standard assignments. What’s often missing from any manual labor scheduling solution are these kinds of tools that enable management to point out special assignment situations.

Building labor standards that reflect your actual requirements is a key part of your focus on skill management. Most importantly, labor standards must reflect your best tribal knowledge about how the most efficient manning for each and every section on your manufacturing floor. When using an automated scheduling solution, your labor standards are built into the system and are automatically put in place with each schedule. As a result, the kind of habitual overmanning highlighted above can be eliminated.

What happens when a supervisor cherry-picks a specific worker for a given position? This is not just a version of the special assignment problem we’ve been discussing? Training is where we need to look for help here. If there is only one worker capable of manning the position, where’s their backup? What happens when they’re on vacation? This points to the need for training management and training planning being tightly integrated with workforce scheduling and not just for new hires.

What are the “Best Practices” (BP) for workforce scheduling? Most managers in operations and HR would like to see these documented. But what is BP?  BP is a method, process, or activity, which conventional wisdom regards as more effective at delivering a particular outcome than any other method, process, etc. when applied to, in this case, labor scheduling. Ideally, BP’s are the best way of resolving scheduling issues for each employee across all shifts, departments or crews. So why even within the same facility is there so little agreement about BP’s for labor scheduling?

In reality, given today’s operating budgets and given the push-and-shove getting product out the door, finding any BP’s written down is rare. You might hear, “It’s all in the labor agreement”. Not likely. Even with a union contract, there are always side agreements that some managers may know about while others have never heard of. There are quite a few problems that workforce scheduling has to solve on a day-to-day basis that are simply not addressed by a typical labor agreement. How to schedule for surges in production or service demand, the opposite, engineering down time, employee job preferences, access to overtime, meetings, requests for time off, training and a lot more that has to be worked into the schedule every day.

When it comes time to undertake a scheduling rules analysis, Tugboat Software is typically first-on-the-beach. For documentation, we typically find a mixture of:
•    HR policies that cover hiring, firing, benefits – a great deal in fact. However they only address labor scheduling in very broad terms. They don’t address the daily volatility.
•    Payroll policies that again cover lots of bookkeeping details and include rules but also don’t address labor scheduling per se.
•    Spread sheets which may track jobs and who’s qualified to perform each.
•    In 24/7 operations something which keeps track who works days on or off
•    As mentioned, if there’s a union, a contract.

When you roll back the stone, what you find governing labor scheduling are “past practices”. “This is how we’ve done it before.” Again, these are mostly unwritten. With an automated labor scheduling project, one of the main business goals is bringing these past practices into the light of day and getting them organized. However when you run a discussion about how labor scheduling actually gets done, it isn’t long before someone says, “I didn’t know we did it that way”. Or, “That’s not the way we do it on our shift (our department)”. Converting best practice into BP is difficult without a lot of design analysis. And a big part of design analysis is getting managers to agree on which of their past practices are actually their BP.

Until you go through the real work, past practices are usually taken as BP. Unfortunately this often disguises, or justifies how scheduling rules are applied differently or unevenly from one department or shift to the next. Fortunately, there are a lot of strong reasons for investing the time and effort into a scheduling rules analysis.
•    Capturing the best institutional knowledge about hot to schedule your entire operation
•    Eliminating the game playing
•   Lowering your cost of labor
•    Reducing unneeded overtime
•    Eliminating grievances
•    Moderating the pressure on your scheduling operation
Then there’s the relief of knowing how to schedule when those skilled managers are absent.

An Automated Labor Scheduling (ALS) solution is only as good as the data it uses and here’s the data we’re going to need. There are five different types of data:

1.    Production Plans
2.    Jobs
3.    Employees
4.    Absences
5.    Skills Inventory

Although there’s more to be covered, Production Plans are addressed in our blog post Production Planning vs Labor Scheduling. The next three types of data should be relatively easy to find and will be addressed in a following blog. However creating a real “skills inventory” usually requires some work and this is the one we’re going to start with. You can’t generate a labor schedule without it.

If the skills data is limited to what’s found in HRIS and/or Payroll systems these won’t include the fine-grain information we need for a labor schedule. As an example, for payroll, an employee may be classified as a packaging machine operator. However which machine is he/she qualified to operate? Is a packaging machine the same on one line as on the next? Operation can vary for an older machine compared to a newer one. Or, the same kind of machine provided from one vendor versus the next. We need to know. If it’s a complex machine, when did they last work this job? Has this skill become non-current? If it’s been some time, can they damage product or are they even safe? This type of data for training and job-assignment history is an essential foundation piece required for a fully ALS solution. So, for each employee here is an outline list of skills data we need:

•    Certification, if required, plus time when certification will expire
•    Qualification
•    When qualified
•    Qualified by whom
•    Last worked date
•    Currency with this qualification
•    Additional training required
•    Any special limitations? lifting, climbing, etc.

In most cases this skills data is “is institutional knowledge”. In other words, it’s known only to those front-line managers who’ve been there long enough to have it at their fingertips. As a result, this data is only available when those department heads or shift supervisors are actually on the job. So installing an ALS solution is a great opportunity to create a skills inventory database which captures this knowledge. Once captured, it’s available for two purposes; when those experienced front-line managers are not around and, when planning and modeling schedules for the future. Also notice that a lot of this skills data is time sensitive. It will expire over time. So a good ALS will automatically capture and update this skills data as schedules are executed and stored. Having the training department or training manager closely engaged with the ALS system so they can add and update skills to the system as workers are trained is a big asset.

The main point: skills data is perhaps the most critical data when an ALS system actually solves a schedule. How many people in your skills Inventory can actually fill each job assignment? If you have to move a worker from one position to another, where will they be most valuable? If they are otherwise unassigned, again where will they be most valuable? These are the kinds of questions that a scheduler has to answer quickly and usually under pressure. This is an ideal chore for an ALS.

Another observation is that we really don’t need to have this data in an integrated payroll or HRIS system. Firstly, the necessary skills data isn’t going to be found there. As you can see from the skills data listed above, this information only really supports an ALS solution. It doesn’t support payroll or benefits. As long as you have a convenient tools for exchanging data with HRIS and/or payroll, all of this data can easily be managed in a well designed ALS solution.

Without this fine-grain skills data, the labor schedule will not be reliable. Without reliability, under the pressure of last-minute changes the schedule can’t be depended upon. Without reliability, the labor schedule wont tell you if your production plans can actually be carried all the way to delivery or the loading dock.

One might think that the longest part of an implementation project for workforce scheduling would be the customization of the software. Our customers at Tugboat have taught us that more often it’s entering local data for job skills into the SOS system that holds up the “go live” date. So this spring, we added a new offering to our services. In February of 2009, I traveled to McComb, Ohio to the Consolidated Biscuit Corporation plant to jump-start that process.

Schedulers and managers worked in advance of my visit to get that information pulled together. Once on site, I was able to enter about 95% of the employee qualifications and job preferences in four days. If the plant had decided to do this work themselves, the project might have stretched into weeks or months, while personnel who already had full-time work squeezed data entry into the nooks and crannies of their workday.

One of the keys to success for this approach was that the plant made it a priority to collect their data in advance, in order to make the best use of my time while there – this helped them set a “data deadline”. Working with their data also raised questions about how the client organized their jobs into categories and departments, giving an early heads up to the project coders who were automating this customer’s employee scheduling solution. Informal conversations with management helped pave the way for understanding of how the implementation would proceed, and for a bit of preliminary training on the system. This was a sort of value-add informed the entire data organization service process in a way that couldn’t really be accomplished by hiring a local temp just to do data entry.

CBC signed on for this project in January, and now in late April we’re in the fine tuning of customization and testing. This might be our fastest workforce scheduling implementation ever, at least in some part because the challenging task of data collection and entry got an early boost from this new service from Tugboat.

Learn more about workforce scheduling solutions here.

Posted by Colleen Fitzpatrick, Project Specialist

Resistance from front line managers when implementing a labor scheduling solution happens for a variety of reason. One reason being that the order has been brought down from corporate. Unfortunately there is always resistance to any software solution that is sent down the pipe from corporate whether it be a good solution or not. Often front line managers feel their current practices are adequate. They do not realize there may be better alternatives and they might not buy into the big picture benefits. They need to get product out the door and don’t have the time to focus on what looks like a lot of work for what to them seems to be a small improvement.

Let’s face it if they can’t see how the solution affects the bottom dollar of their labor budget right away they see no point in focusing on a new project. This can lead to a setback for the project and can contribute to a downfall for the facility. This is unfortunate because the long term benefit – controlling labor cost – far out ways the short term impact on front-line managers.

Short-Term Benefits

Besides, there are other very short-term benefits. Schedulers have more time to focus on operations and other management responsibilities. Getting the right qualified worker on the right job is much simpler. The labor scheduling is done according to the rules that the facility has agreed on. Managers and production planners have the ability to make last minute changes and create “what if scenarios”. Analysis of your labor utilization can be done with reports. All of this is impossible to do when you are managing labor scheduling by hand. Altogether, the solution saves time for H.R., payroll, supervisors, and schedulers. It is not easy to put a dollar value on these kinds of savings and flexibility.

Posted by Jennifer Holmes, Implementation Specialist

We recently completed work with a customer where we were able to reduce their overall cost of labor and measure the results as well. The goal was to reduce the use of temporary workers (Temps).

Temps are an ideal resource when balancing the ups-&-downs of production demand with a finite labor capacity. Using Temps is more cost-effective than having to fill open jobs with overtime. So eliminating the use of Temps altogether is not practical. However, relying on Temps indicates that one’s existing work force is not being used in an optimal manner. And this is an indirect indicator of labor waste. If you can meet your production demand by using your existing work force in an optimal manner, the use of Temps is reduced and your overall labor cost goes down.

How? We used optimization to find the best possible match between the skills inventory available with the customer’s hourly workers and, on the other side, the labor demand required to meet their production requirements.

Tugboat’s customer operates a prepared-meals manufacturing facility with about 800 regular hourly workers. They were using 20 to 40 Temps every day. Using Tugboat’s scheduling engine to do the crewing automatically the number of Temps was cut to 10 a day.

Measuring Labor Waste

In this case, the cost of labor waste is very easy to measure. Before Tugboat’s solution 20 to 40 Temps. After, 10. Temps may not sound expensive, but. The savings calculates out to at least $207,000 per year. (How would you like to take part of that home as a bonus?)

Here’s how we arrived at that number. Using the conservative estimate for reducing Temps from 20 to 10, that’s 10 fewer Temps. At $10.35 per hour, the cost of a Temp comes to $82.80 for an 8 hour work day. Multiply by 10 Temps, 5 days a week, and 50 weeks and we get to $207,000 per year. And, this should be easy to verify. Finance can report the cost of their temporary labor force from any prior period and compare it to the costs for the current period.

Posted by Terry Schilling, VP Marketing & Sales

In a recent article in the Manufacturing Business Technology magazine, Julie Fraser discusses how “software can bring some capabilities to the plant environment that are very difficult – if not impossible – to achieve by other means.” In particular, she highlights resource allocation as a need that can only truly be addressed by systems that look at multiple products and a highly skilled workforce that cannot all be cross-trained. Software also provides visibility into the plant through automated data collection, and performance reporting. The article, “Why buy plant software in a Lean environment?” can be found HERE.

Posted by Terry Schilling, VP Marketing & Sales