Many years ago, I proposed that machine control would change the surveyor’s role and made it a point to discuss this with many state surveying groups. As a rule, surveyors need to be exact. A circumstance could come up where one would have to defend their work in court. With the need for accuracy, surveyors were not happy with contractors bringing precision equipment onto a site. Some thought contractors were cutting costs by reducing survey crew trips to construction sites. However, the real reason was that a surveyor may have to go back and tell a contractor some of their work was done incorrectly. Every surveyor hates to return and re-pound stakes that got knocked over, so how do they make sure the contractor can perform great work independently?

Survey Builds Data

I have been involved in lively discussions to legal debates over the production of machine control data. Regardless of the state, the answer is always the same: data built by or on behalf of a contractor is not an issue. Once that got settled the question remained, “Who should build the surface you will grade to?”

Traditional Surveying

Survey is solely about points. Load calc points into a data collector and go out and place them on the ground. The contractor connects the dots and a surface is made from the points.

The image at the right shows what we were given in a set of plans. Points found on the plans are placed on the ground in three-dimensions with a stake. The cut/fill to the desired location was marked and the dirt was moved. When things got close, blue tops were placed with the top of the wood stake at finish. Other types have a staple in the top holding the whiskers, a good operator would “polish” the staple and get grade that precise. The difficulty in this method of staking/grading is what happens between the hubs, a low spot or incorrect drainage is easy to miss. Only when a lot was paved would the problems show up. No matter how much spot checking we did, sometimes a spot would get missed.

Data Models

When I started doing this work, the only frame of reference I had was how things were staked and cut in the field. Because first generation equipment was difficult to use and problematic, training and ongoing support were normal. I spent more time in the field than in the office. We would build the model and I would be in the field for days, to weeks, training crews and getting a better idea of how to leverage the data that I collected.

The knowledge I gained by working in the field allowed me to create workflows that are now commonplace. The other problem we faced was that nobody knew how to build a model, but surveyors felt they were the best option. This is a learned ability and with practice anyone can become proficient, like me for instance. I worked long and hard to figure out how a model works in the office and the field. Coupled with a good understanding of how the mouse clicks transferred to the ground, I began to connect those dots. One of the most frequently asked questions I get is who the best candidate for a data builder is. I find it easier to train a computer savvy field person than try to get a CAD expert to think in three-dimensions.

In time, more people learned how to create good data and software ran ahead to improve commands and performance. The future holds the key to 3D design and implementation of data into the field. BIM has taken the lead due to multiple trades trying to occupy the same physical space. Currently much of that design work is done with CAD technicians producing the 3D models suitable for construction.

The ability to design, (or produce 3D data from CAD and plans) is mature. We will discuss the integration of surveying and data.

Cross Training

As more contractors started to create their own files, surveyors were not building as much data but keeping busy doing the important job of positioning. It’s necessary to understand what a surveyor does, but not actually do it. Surveyors need to understand data but not necessarily know how to build surfaces. Here is what each needs to know about the other.

Somebody must make a model for the machines to work. With model in hand, it’s time to go to the field. Surveyors are critical to any job large or small. I have seen the start of too many jobs by well-meaning contractors placed in the wrong spot or incorrect elevation. The nuances of positioning are complicated, and it is the surveyor who can assist in correct site placement. The GPS lead on the jobsite needs to have some of this ability too. The most important thing is sensing an issue and when to contact survey for work to be done. I have a surveyor friend who ten years ago would build the model and give the contractor a rover with the data loaded. They would start the job and call when they needed something.

I need to be clear; office and field civil construction workers are not doing survey. We are laying out and grading to information the surveyor knows is in the right place and will perform as intended. That is the end point of data and survey working together, so how do we get there?

Data Prep

I’ll run through the process when we work with a client on a typical site job.

• We build the initial model and do the following:
  • Fix obvious errors and note them.
  • Make grading changes to reflect proper water flow.
  • Verify dates and changes to the plans to confirm current versions.
  • Report these changes to the contact(s). Usually this is the contractor. As the jobs get larger, we are often asked to communicate directly with the engineers and surveyors and include the contractor in the discussions.
• Responses are received regarding questions.
• Data is updated and work continues.

The data needs to start out as a representation of the finished product. There may be changes pending but not so detailed that we must redo a lot of dirt work. This rough grade file will keep the schedule moving and let the engineers know how long we have until we run out of things to do. Hard deadlines go both ways, stakeholders need to know we are faster at moving dirt and need answers to questions.

Your surveyor is invaluable at this stage, coordinate values need to be correct, the best way to know this is to bring them in early. As hired guns, we communicate with survey crews on jobs all the time. Be sure to do this early and with all jobs.

The next step in confirming data is to bring in control and verify locations and accuracy. Now that things are in the right spot and the data is close to correct, we can send the data to involved parties and gather input. Don’t expect a lot of information, everybody is busy, and most don’t have the time to review your data. It’s more of a courtesy.

Survey Functions

I have often been involved in some very detailed discussions about survey’s role on a job. I am an expert witness regarding disputes that sometimes involve surveying. I am no surveyor but have a good grasp on the practice. It takes time to be good at this and I leave it to the professionals.

The surveyor needs a plan for what they will do on your site. This prevents duplication of effort as well as ensuring it is going correctly. Here is a list of things to get right.

• Understand what a site calibration (localization) is and what the report means. These numbers are critical and you need to understand them. When the surveyor is using different equipment than you are, you will need to perform this on your own. Compare the results to theirs to verify.
• Know the difference in collection times. Sometimes you can walk and click, and other times you set up the bipod and cook for a minute. This affects accuracy and a few more seconds on various point types will be rewarded.
• Respect procedure. Don’t take phone calls or talk to people on site when you are involved in critical tasks. When the surveyor tells you not to bother them now, it’s not because they are a jerk. They are in the zone and don’t want to miss something. When you are laying out curb points for example, always do things the same way. For example; I always do PC, RP, PT. Pick your method and stick with it. Here is a simple curb layout. This is what I want to see when I’m laying out.
• Perform daily check-ins. Know that when you start your day you are as correct as you were yesterday. Do this for rovers and machines.
• Share and listen. The information needs to go to the surveyors and engineers for review and comment. Again, you will not hear much, it’s just good to let them know what things look like in the field. This is a big advantage on remote jobs.

To wrap things up, learn enough about the other person’s job for better communication and efficiency. None of us can do a job that we are not trained for, but an understanding of the roles around you will go a long way in making things run smoother.

The request for “a quick takeoff” means different things to different clients. When a client makes this request, I generally know exactly what they need. The quick dirt number I provide usually leads them to knowing how much Teflon tape they’ll need for the water pipe joints.

The procedural filters we use while doing a scope of work will change over time, coinciding with the different stages, to make our job easier and more productive. Clients each have different requirements for their takeoff, eventual bidding, and final production. To walk you through this process, I will begin with the basics and investigate advanced ideas while I progress.

During this offering, I will speak in the first person, like it is “our job,” acting as a consultant who will perform the takeoff and processing services for our clients and not perform the work ourselves.

Just a Takeoff

When we “old timers” used to receive a request for a takeoff, the rules were simple and worked well for years. With more technology introduced, the deliverables clients expect to receive are much more advanced. Acquiring additional information, in respect to a civil site takeoff, may help you get your numbers right. The data in hand is purely for bidding purposes, but clients often want more information than they need. It only wastes their money.

Takeoff to Project Management

Once a client has won a job, I am all for details and more information. Many of you know my old line, “if you’re using your takeoff for data, you are doing too good a takeoff.” The same holds true for takeoff detail. If your takeoff can instantly become a project management document, you’re wasting your time with a too detailed takeoff.

The difficulty comes from transitioning a file from takeoff to production. Many think that you need to start over in order to make things work for production. As an old timer, I would agree with this, but industry software has made this a non-issue. Here are some of the transitions we need to make when the job is won.

The Dirt Number

As an experienced estimator and large-scale project manager, I always keep some money safely tucked away in different scopes for the eventual rainy day. The dirt number was one way to do this. During the bid phase we might have listed the strata from bore logs, but often they are not available, or time won’t allow it. Now that the job is ours, I start the deep dive into the actual cost of dirt moving and investigate the following items:

• I get our drone, or hire a local drone operator, to fly a pre-start topo. We all know once you mobilize to a site and put a tooth in the ground, all bets are off for another trip to the well for more money because the OG information supplied in the plans was incorrect.
  When we get the topo information, the takeoff is rerun. If we are better than before, I use the advantage in my rainy-day fund. If the numbers go bad, we call a meeting with the owner and renegotiate. Make sure this is all done before digging. When fast-tracking, you still have the date the topo was flown as well as images that show no disturbance. An email to let everyone know this is an issue will serve to keep the issue open until a change order is processed.
• The different amounts of each type of dirt that needs to be moved is the next area I look at. If you had the bore logs and they were entered into the takeoff numbers, review them and start to look at actual costs for handling each material type. Many contractors know their areas well and will assign an average dirt number to the quantities at bid time and come back later for refining. You can now figure amounts much more closely with additional information.
• For mass haul analysis, cut/fill and enter our average number for dirt moving works well for bidding. These quick numbers are a result of careful figuring based on prior work and should get you in the ballpark to win some of the hard-bid jobs you go after.
• When you get the job, it’s time to start drilling down into the numbers and squeeze those few percentage points to make the boss some money. The amount of dirt getting moved at what distance on what quality surface is the breakdown. I will review a site mass haul in the video and go over what I look for in the reporting.
• Many site jobs require going offsite to dispose or import material. This is another chance to make some calls, shorten distances and lower costs. During the bid, you might have used local numbers or made a quick call to plug in a price. Now is the time to get some savings from the averages used in the proposal.
• Now that we have won the job and returned from the celebration, it’s time to be good to the owners, and help us a bit along the way. We all need to bid to the plans and specifications and consider the pricing on additions and alternates. There are too many variables to go over here, but each contractor knows that there are a lot of better ways to do certain parts of a job than what’s been drawn by the engineers. We see the biggest disparities in chain business when plans have been drawn out of state without intimate knowledge of the area. Others have some luck by having worked in the immediate area and are then able to recommend some changes that will enhance the job and save everyone some money.

To this point, I have used quick and basic takeoffs that drove the dirt numbers without having to redo anything. I just spent more time drilling down in the listed areas. I made a lot of owner’s good money by being diligent in the above areas except when back in my day we walked the topo, drones weren’t invented then.

You need to perform the above first. This ensures the original ground information is good, you understand conditions, as well as knowing how far you’ll need to go for material import or export. With this completed, it’s on to the next phase.

Project Management Process

A lot of time is wasted moving materials on a jobsite. It could be dirt that was set incorrectly. To pipe and other import items that always seem to be in the way. We now need to elevate the quality of our data from a takeoff to a performing site data model. The big question is whether to start over or improve the takeoff to data quality. Here is my process:

The Big Stuff

My first question is, did the overall footprint change? With the basic layout still intact, the improvement of the takeoff is my first choice for a data model. We already have the layers broken out and most of the COGO (Coordinate Geometry) is good.

Many times, there are changes to the plans after we have won the bid. Hopefully some of them are from our value engineering proposals to the owners and engineers. With that information in hand, a fresh look at the model will tell you if you should start over or improve things that were in the takeoff. With good layer naming and consolidation, it is not hard to add islands or redo a changed curb line. Resized retentions and other common areas are easy to remove and replace. I will go over a few of these points in a video regarding the use of layers and how to use them to your benefit.

The Details

When we do a takeoff, little attention is paid to making a parking lot look great. We elevate curb lines and possibly change bad spikes. I have done countless studies and presentations from a takeoff to data model quantities and the difference is miniscule. The price of a data model is about 3-times what a takeoff costs. It is not worth our client’s money to get crazy detailed with a takeoff. During the bidding phase, plans are often not approved by all agencies and will change before being issued with final approval. It is important to note the delta changes on the latest and greatest set. A word of advice; never trust clouds on the plans to indicate all the changes made to the job. I like Bluebeam sheet compare for this, I will do a video to explain.

With a final set of approved for construction plans in hand, we can get to work. Let’s elevate the takeoff data and prepare it for construction.

• Drill down into the dirt as explained above. Get those numbers correct and detailed.
• Break out the job the way you need for phasing and ordering. You have gross numbers, now you can meet with project managers and superintendents to organize things better.
• Verify the data model is good, (at least for now) with current changes. During data building, we will find issues and submit questions. In the mean time, you can leave the data as-is for the rough grade phase. If the questions would affect the initial mass earthwork calculations, leave the area blank in the model until you get guidance. Don’t move dirt twice to look busy.
• Add numbers to the quantities. All current takeoff software can export to your spreadsheet estimating program or Excel for adding prices.
• Go back to the takeoff and add any additional information you want to run through the process of measure, export, and price. We will do everything from count light poles, measure striping, breakout straight and curved curb, and place playground equipment bases. With the plans in front of you digitally, we find it good to even measure items that will be subbed out. It sure is nice to adjust a supplier’s estimate and save some money.

The U.S. Interstate Highway system is almost complete. Regional networks are mature, and the new right of way is geared for housing. Luckily, we still see new alignments and the percentage is increasing for repaving and full reconstruction. This increase has led us to become efficient with sometimes difficult road improvement jobs. Let me explain.

Types of Rehab

There are three basic types of road rehabilitation all requiring a different approach to the data. While I have defined the types for data purposes, there are projects that can be a combination requiring a change in the process.

Full Rehab

This is the easiest and currently the most popular type we perform. The road may move horizontally and/or vertically and none of the previous roadway elements are to remain. Paving, curbs, and driveways are all replaced. Ditches will usually be reworked, if there is an underground storm system, it may change as well.

With everything being new, it may not make sense to refer to this as rehab. The reason is at some point the work you are doing will need to connect to something that will not be moving. Driveways, buildings and off-right-of-way drainage are required to match-in with the least amount of disturbance possible.

Mill and Fill

Asphalt paving does not have a great life in many parts of the country. Freeze-thaw in the North, water and heat in the South, and brutal sun in the West define the finite life and eventual repair of the wear surface. When spot repair is no longer possible, the road surface will be milled, and a new full-depth mat will be laid on the freshly compacted base.

We will go into detail later, but the curbs are usually in good enough shape to keep. We are now required to respect the vertical and horizontal constraints of the existing road. In addition, there are required minimum and maximum coverage and lifts for the base and wear surface. This gets difficult to work with, as there are many constraints plus having little or no ability to try and fit everything in.

Widening

Right of way acquisition is usually done with the future in mind, especially with larger arterial roads. As traffic counts increase, lanes are added to accept the load. When confronted with a widening job, we are concerned with two major points. First, the condition of the edge we are joining to, and then the topography of the extension. I will detail these in a moment.

The Basics

When doing rehab work, we usually go the route of a road job. Alignments and templates are a good start to get things right and, most importantly, a way to easily adjust as things change. The single biggest issue with rehab projects is the dynamic nature of all the parts that need to come together. Here is an outline of what we like to see.

Horizontal Alignments

Ordinary COGO (Coordinate Geometry) will be provided when doing a new road. Sometimes in a rehab job, the plans call out for following the existing road center. Yes, this is a loaded statement from the designers. There may be a line on the plans that might be an old alignment, or something drawn for convenience. Alignments with good instructions are easy to get on your screen, but what about that mystery scratch in the near middle of the road? There are several alternatives.

If you can’t get a good centerline from the CAD, you need to decide on how to guide machines for the work. A best fit centerline takes the edges of pavement and averages the distance between them to give you something to steer to. Here is a set of shots that were taken along the existing edge of pavement along the slip formed curb/gutter.

There are many ways to get an alignment, in this case I am using Carlson’s Best-Fit Centerline. You can use points or lines. Here I drew lines through the points with arcs in order to give the program a smoother start to figuring out a centerline. The alignment is drawn through the upper line (shown above) and will offset to get it to the approximate center.

I often times will do this on the other side of the road as well. When this is necessary, you need to either make a new averaged centerline or create two different roads using one for each side. These cases are generally for older, small streets and roads where environmental conditions have caused heaving and erosion to move the curb. Logic says we will replace those sections so this is a rare exception, not the rule. Knowing that, you are now able to fix real issues.

The new centerline is shown here. It deviates from the edge by almost 3-feet in the worst spot. This is too much. Changing the parameters will tighten things up. This is shown in the video.

With a good alignment to steer to, we will now work on the vertical.

Vertical Alignments

With horizontal alignments we are trying to get a centerline close to actual. With a vertical alignment, the stakes are higher. When the vertical profile corresponds with the centerline, it must follow road speed rules regarding cross-slope and the finished job needs to look good.

The job depicted above has 3D shots along the edge of pavement, as well as, centerline shots that are turned off. With these 3-points acting as cross-sections, we can create a good existing road surface as cross-sections. Don’t be alarmed if a contoured surface looks bad, a cross-section look is the best way to generate a finished product.

I have used Carlson’s Road Rehabilitation Profile command with good results. You will need to get things in order, including some of the outlined steps I will reference. When the pieces are in place, you will have access to the options shown in the dialog box. Lots of power is provided to automate a sometimes-difficult task.

With the alignments done be assured you will be revisiting these to make things work better. We will generate cross-sections to verify the quality of the data.

The following is the Carlson section alignment dialog box. There are enough options to give you the results you are looking for. I am using these options for the job in this article.

With the section interval and special stations defined, I will now create sections of the existing road to give me an idea of what I’m looking at.

Gathering Existing Data

I need to mention a critical point. When we work with clients who build models, we insist the shots are taken with a total station. GPS accuracy is not reliable for the number of required shots and would take too long to get low residuals. A robot and one person can shoot quickly and with accuracy to make sure we are not wasting time. We accept GPS shots only to fully rebuild the job after receiving good data.

Section Review

The production of cross-sections will be the true test of what needs to be done next. I have spent a lot of hours getting things ready for production on rehab jobs. The constraints of the existing features that remain and the rules imposed by the parameters of the road design make the task difficult. It may be necessary to rebuild the job several times in order to make everything work.

Also be aware of the hierarchy of importance in case something must give. For example, we need to keep the curb but need to go less than the required 2% slope. That’s an easy one, but there may be situations where several rules need to be flexed for things to work.

Here is one such example. The road is straight and calls for a 1.5% cross-slope. The right side of the road is almost flat. We need to review up and down station to see how far this extends. The solution here was a 75-foot curb replacement due to heaving of the existing curb and gutter.

With requirements that are often out of the requested values, this type of work takes more time. Many rehab jobs just want to follow the existing road, mill the existing surface and come back with minimum cover. This type of job still needs data to work correctly.

Carlson has a tool I have used with great success. The Match Reference Section Slope command allows you to specify the desired slopes and the limits of deviation to it. Here we are trying to get a 2% cross-slope with a variance to try and make things fit better.

After filling out the dialog box, the command has listed the varying cross-slopes generated by the settings.

Adjusting the Parameters

There are two tools I use to verify a rehab design. The first is the actual material to be used. At some point there was a takeoff done, and I want to make sure we are in the ballpark.

This is just a portion of the report, but the totals are in line. To adjust things, go back to the Process Options dialogue box and check the Adjust Template Grade Table. The side not associated with the Profile Grade will adjust. This may push the slopes outside the design parameters and require a variance to get the volumes down.

I will then plot the sections to verify the profile is doing what I want it to. With a small road like this there is not a lot of room to move, but if the job is several miles long small tweaks can bring big savings.  This particular road is getting 8-inches of white paving. The red is the sub-base needed to bring this up to grade. Had this trend continued for several stations in both directions, I would revisit the vertical profile to try and pick up some material savings.

Summary

The task may sound daunting, but the job needs to be broken down into the individual parts that make up the job. I have tried what seemed to be quicker and easier methods, but changes are near impossible and always take longer.

Approach each part of the process as a separate task and the delineation makes things easier to imagine. With more of these jobs coming along every day, it pays to be proficient.

I have featured Carlson because I have experience using the commands for road rehab projects. Other software can accomplish the same tasks. The commands will be different, but the procedure remains the same.

More contractors are taking advantage of stringless curb technology today. At TOPS, we got involved with stringless curb when it was proposed as an alternate application for machine control. In their quest for increased productivity, some of our clients are early adopters which we credit with our first-hand experience in using the technology.

The Concept

The idea is straightforward. Adapt the use of alignments for paving systems to a curb machine and eliminate the string. There isn’t more to do for the transfer of the guidance and the technique seems easy to perform. Unfortunately, we encountered some problems along the way. But lucky for you, we have taken the time to point them out and guide you on this process.

Stumbling Blocks

You need to build an alignment-based project that has the usual elements, horizontal and vertical alignments, and a template. Brands vary but the template can be used to pick the side of the alignment the curb is on, as well as, slope for fill or spill curb.

One of the difficult things to do is to make the alignments as if you’re in the field doing the work. This varies for our clients and it is something you need to coordinate with the curb crew.

In the example here, I bring up two interesting points:

1. This job is the addition of more parking to an existing lot. When joining to that curb, we need to get accurate shots in order to smoothly pick up the slope of the current to future curb.
2. The 90-degree corner will necessitate a stop in the alignment. The question for the field is where to start one alignment and stop another.

Alignments

As with any road job, a horizontal and vertical alignment are required. With curb, things need to be different. There will be some figuring with both.

Horizontal

Each curb is a separate road job. As shown earlier, you need to coordinate with the field as to the start and stop points. Experience has led us to have this consistent, but it takes time to coordinate. When you get it figured out, it will stay standard for the most part.

When working with a closed island with all curves, the line needs to either stop short of the end or go past it and not be on the same path. This example shows the alignment stopping .10 feet from the start. This gives the machine a chance to complete the run without the software problems of an alignment running back on itself.

We will also have the alignment bypass the start by a couple hundredths of a foot when it gets back to the start point to keep the lines from intersecting.

While building the alignments you are also providing a full takeoff of the curb so the field can schedule concrete and plan the pour accurately.

Vertical

This is where things get interesting. We all know that good plans have elevation callouts for the major points of a curb. This example is trying to do that, but this job has sheet graded contours that make things more difficult. We need to pay attention to closed areas that may trap water and make them back-flow into the main slope plane.

Breakover Angles

This is by far the most critical part of designing stringless curb files.

Here are the elevation points as called out in the plans. The curb moves along but there are angle breaks with a 2% delta. When entered into the machine like this they will cause it to abruptly change slope and make a mess. To remedy this, two things must happen.

1. Vertical curves must be added to the alignment to smooth out the transition in slopes. Our method for figuring the amount of curvature has been derived through experience working with machine control and curb machine vendors. Experience will need to be your teacher here.
2. After making the curb look right, the new edge of pavement 3D information needs to be incorporated into the model so the subgrade and paving are not affected.

Here is the alignment after the addition of the vertical curves. The transition is now smooth, and the machine will make the slope change gradually so things look right and perform well.

In the images, the difference may seem subtle. In the field, it is scary to see the machine try to do an instant 3- or 4-degree slope change. I’m sure the question will come up in a cart and horse fashion. If you are changing the parking lot surface, should you design the curb first? Most jobs don’t use stringless curb. The ones that do are usually requested after the surface file has been made and the curb contractor wants to use the technology.

The initial file creation is procedural and a process should be followed. This is because there may be 50 alignments for a big site job, and you don’t want to go back and check every line to see if you missed something. There isn’t much rework involved after the curb alignments are returned to the surface, we just want to make sure base depths conform to plan. The following is an outline of the entire process.

Create Curb Template Alignments

• Save the file as a new version to keep the surface file alone
• Create breaks for machine control
• Create a vertical and horizontal alignment from the curb lines
• Station the horizontal
• Add vertical curves to smooth out the profiles
• Create the proper exportable road file. Brands and requirements vary

Create New Surface

• Do another save with a new name
• Remove the use of the initial curb elevations in the model
• Set the vertical profiles as the new curb lines
• Offset the lines in three-dimensions to get the lines locations and elevations to gutter or edge of pavement
• Adjust the surface in areas that make it smooth and well drained

A bit easier said than done, but experience has really helped us get this operation efficient. The curb files we make for clients can guide stringless curb with confidence. I remember years ago when the head of Gomaco asked me why I thought anyone would use stringless and how I planned on giving the crews confidence to spend tons of money and time with no string to lead the way. Here is what I stated and how a company gains trust.

• The curb (or white paving) is derived from the model used to blade the surface. If that looks good, things are okay.
• Use a 4-wheeler to do a dry run. Load the road job and drive along the curb, you will see any problems before the pour.

With all these advantages to stringless mentioned, file preparation is not a big chore. I have some sad stories about stringline that caused us problems over the years. A few dry runs of practice and maybe some “air-paving” will get you comfortable and ready to make the move to automating curbing and paving.