The Thorny Problem of LiDAR Specifications

By Robert A. Fowler

When deciding to contract LiDAR most people have some concerns because the number one problem seems to be there are no legitimized specifications which have been written to address this technology. Well, that’s not quite true. FEMA has some draft specifications which, in the version I have at least, are impossible for anyone to comply with unless you bulldoze the area devoid of vegetation. The American Society of Photogrammetry and Remote Sensing has a LiDAR committee, which is looking into drafting specifications. These may take some time to produce as there are a large number of stakeholders, so in the interim it plans to issue guidelines.

The problem is there are a number of ways of looking at this subject. To produce specifications that are fully conversant with the different technologies already on the market, almost requires a specification for each type of laser system.

To understand the complexities of the issue, it is necessary to consider that LiDAR systems are three independent bits of technology which have been cobbled together by different manufacturers in slightly different ways and with different types of components.

There is the basic airborne GPS (Global Positioning System) which is made by several different manufacturers. You might think that with the state of the art today these all work in the same way. And in theory they do. However, they do not all deal with timing issues in exactly the same way, and they use different software to process the satellite data before outputting it into coordinates you can use. Consequently, there are small differences in how they collect and process the time issue. Now on the ground, on a stationary control point location this doesn’t make any difference, and in a kinematic mode in a relatively slow moving ground vehicle it also doesn’t make much difference. However, in an aircraft flying along at 60 meters a second or faster, a timing difference of a few hundredths of a second can make a difference of several meters in position.

The second piece of equipment is the inertial measurement unit (IMU). These come from a variety of manufacturers as well. Some manufacturers make various qualities of IMU. As a consequence, depending on how much you are willing to pay, the IMU will be more or less accurate in determining the position and the angles it measures. Again these are critical functions tied to a high precision clock. To give you an idea of the variety, you can pay anywhere from a few thousand dollars to a million or more for an IMU. Optech, one of the major LiDAR manufacturers, uses a system that is worth close to a quarter of a million dollars. The system has a response time and an accuracy that they believe is adequate for most of their clients, within a reasonable cost range.

It’s reasonable to assume that if the system proposed by someone else had a $25,000 IMU, it is unlikely to have the reliability, repeatability, response time and therefore accuracy of a system ten times its cost.

Finally, there is the laser unit itself. There are a vast variety of lasers being manufactured for a wealth of different purposes from medicinal scalpels to torches for cutting steel. There are, surprisingly, a large number of measurement lasers and several scanning lasers on the market. These basically, send out a series of light pulses (usually in the infrared end of the spectrum) which hit an object and the return echo is received by a light-sensing device. The scanning lasers also come in a couple of varieties.

The light emitting diode which is the light source, can be imagined as something like a light bulb. Swinging that from side to side multiple times a second is going to shake the LED to bits in short order, so in many scanning lasers a mirror swings back and forth instead, directing the laser beam through a swath from left to right. Most airborne laser scanners use a moving mirror to effect their scan. The disadvantage of this system is the mechanical aspect of having a high speed swing back and forth. As the mirror reaches the end of its swing, it perceptibly slows before stopping, reverses direction and builds up speed again for the return trip. Because the mirror is swinging many times a second, this slowing down and speeding up, all calculated in microseconds, alters the possible accuracy of the pointing.

There is theoretically a way around this. The manufacturers could use a rotating prism, which just turns one way. Then the speed at which the prism is rotating can be kept constant. Indeed, some manufacturers of scanning LiDARs use this type of technology. The disadvantage of the prism method is in timing the precise angles of the swath being collected. There’s no fixed angle stop position to correlate to a time signal. The other disadvantage is all of the data are being collected in one direction. If there are any biases in the measurements, they will never show up except with a field check.

While there are good arguments from both sides, the current preponderance of opinion in the field is, the mirror type system, despite its drawbacks, in fact does produce a more accurate result. The way around the slowing down and speeding up is in the software, and in ignoring the data collected at the very end of each swath. Most LiDAR service providers do that, which is why they may tell you they collect a swath 3000 feet wide, but they only use the central 2500 feet.

So, we have these three bits of equipment, all doing their own thing, collecting data tied to high precision clocks. The LiDAR manufacturers have to ensure that the data collected by all three systems is in sync. They have to be able to calculate the position of the GPS and the IMU for each pulse of light from the laser.

This is where the second timing problem comes into play. Most airborne GPS units capture position at one-second epochs; that is they record a location using four or more satellite signals every one second. However, as we stated earlier, an aircraft flying at 130 miles an hour is travelling close to 200 feet a second. During that time the laser scanner is going back and forth quite a few times. Likewise for the IMU data, the all-important attitude, or angles off vertical the aircraft is in at any one point are typically recorded every half or one second. That means positions and angles are interpolated for each single laser pulse at any one instant in time.

If this sounds pretty convoluted, it is and it isn’t. There’s a lot of stuff all going on while the LiDAR is working, but broken down into small component parts, manufacturers can develop software to calculate all of the information needed (including the slowing down and speeding up of the mirror problem) to produce the right answer.

Now you may be saying so what to all of this. But if you are going to determine real LiDAR specifications a lot of this information and how it gets put together is important, because it materially affects how good a result you may expect to achieve from any particular model at any given elevation above ground.

But, of course, it doesn’t end there. Let us assume we have information collected from a system which is airborne, somewhere above the ground. What is really needed is to be sure that the frame of reference of the airborne system is the same as the frame of reference on the ground.

That brings into play a whole new dimension, if I may be permitted a pun. And, again, it is not as simple as it sounds. Obviously the first consideration is the GPS unit. In order to ensure the airborne GPS unit is working in the same coordinate system as the client on the ground (and the whole purposes of the effort is to produce ground information), we need a common reference point. That is achieved by having a second GPS as a ground station operating at the same time and using the same satellites as the airborne system. Because of satellite configurations, and features on the ground that might get in the way of satellite signals (such as trees or tall buildings) most LiDAR operators want the ground station to be located within the project boundaries. If the project is large, then a ground station every twenty miles might be needed for high accuracy surveys.

The reason for this is: as satellites orbit some will drop out of view and new ones will appear above the horizon. The operator in the air wants to be sure he is receiving the same five or six satellites as the ground station, otherwise the translocation solution will not be as robust. If the ground station is a long distance from the aircraft, the chances are very good it will not be receiving all of the same satellites.

Secondly there is the pesky problem of the shape of the earth. Satellites orbit the earth on an orbit based on the center of mass of the earth. Unfortunately, that is not the real center of the earth’s shape. Not only that, the earth is not the regular shape it appears to be on a map projection anyway. So there are differences among the “perfect” orbit of the satellite, the “perfect” shape of the earth for whatever projection you are using, and the real hard ground you might be standing on. This results in what the experts call a geoid/spheroid separation – where the geoid is the real shape of the ground, and the spheroid is it’s theoretical shape (in the map projection). And, if you want to think of it this way, the orbits of the satellites are based on another spheroid totally disconnected from the other two! This separation can vary measurably over distances as close as forty or fifty miles apart. That’s why ground control is needed fairly closely spaced when you are producing very high accuracy surveys.

There are some other fundamental effects that can occur with airborne GPS surveys which can also result in a shift. There’s speculation on what causes this, but some of it is likely atmospheric conditions causing bends in radio signals. These again cannot be determined except by having ground stations using the same satellites as the airborne system on the project site. The reasoning is, any distortions affecting the airborne system also affect the ground system. As the ground and airborne systems are closer to each other than they are to the satellites, corrections can be applied effectively canceling these errors.

But, if you haven’t given up by now, there’s more. Now that we assume we have sorted out the ground/airborne GPS problems, we still have that IMU to check for. How do we know that when it says the laser unit is sitting in an absolutely vertical position that the laser really is in that position?

Well, the simple answer to that is once everything is bolted together we can bench test the equipment. But, you might say, bench testing and running the system several hundred meters up in an airplane are two different things. And that is correct. So once a ground based calibration has been completed, we need a test area where we know the position of the ground features precisely, so we can fly the LiDAR over it and collect a data set. If we fly the system back over the site in the opposite direction, and next switch the flight lines by 90 degrees and do the back and forth test again, we should get identical results on the test area. If we don’t and there is a consistent variation between the four different flight directions, then this is most likely a bias which can be removed mathematically. If there are consistent vertical differences, these can be assumed to be a vertical bias, and again removed mathematically. While this may sound a little hit and miss, it really isn’t. Practically all electronic measuring equipment, whether it’s a total station, or a rangefinder or, for that matter, a dumpy level should be tested against known baselines for biases from time to time, and this is essentially what is being done here. What the technicians are looking for is consistency. However, if the results are inconsistent that means something is not working properly and more bench tests are required.

Essentially, this is what almost all LiDAR service providers do. On a periodic basis, they test their system against a swath of ground data which has been extensively surveyed on the ground using alternate methodology.

So how can we write specifications which cover all of these diverse characteristics and phenomena, but not really have to write up specifications which may be limited to one unit or type of unit?

Let’s put the specifications problem in context. When we buy a car, we don’t ask the salesperson what tolerances were used when the engine was put together, or whether the paint is a specific formulation. We ask ourselves and the sales person about the things we are really interested in. Is this car from a recognized manufacturer that I think I can trust? Will it do the things I want it to do? (And this is based on such things as will it hold two kids, the spouse and camping gear? How many miles per gallon will it get? And so on.) And finally, of course, will it be delivered when I want it and, very importantly for most of us, will it fit my budget?

When we think of LiDAR specifications, the same sorts of real questions need to be asked, summed up as in: What do I, the customer, want?

In the case of LiDAR the answers must relate to the quantity and quality of data so the follow-up questions become: How accurate do I want my data to be and how much data do I want?

Let’s first consider the latter point, about volume of data. This depends very much on what the use of the data will be. One thing that almost every new LiDAR customer is surprised about is the volume of data generated. Realizing you have a point every ten feet means about 279,000 points per square mile.

That doesn’t sound too bad, but don’t forget that each of those points could be composed of eleven digits plus one decimal point northing, ten digits and a decimal point easting and probably seven digits and a decimal point elevation - that's a lot of bytes of data. The raw data files from the LiDAR unit are even larger.

Then we get into the more rhetorical questions such: as a client do I really need the first and last and intermediate returns? If so, ask yourself why? Much is made of these by LiDAR manufacturers, and there are legitimate uses for them, but the average client probably doesn’t need them. In open areas the ground is the first thing the LiDAR beam hits – it is therefore simultaneously the first and last return. Likewise, if the beam hits a building the first and last return are synonymous.

If it hits a tree, then depending on the system and the size of the beam and what precisely the beam hits (a single leaf, a twig, a branch or the center of the trunk), it could register a first return, several returns from branches on the way down and maybe a last return from the ground - if there’s enough signal strength left by the time it gets there. Now, truthfully, there isn’t much value in all of those data unless you are a research forester. Further, if you think the amount of data you get with one return is a lot, wait until you see the size of those data files for multiple returns!

Let’s assume you are not a researcher, in which case there are really only three types of client requirements for land based LiDAR data. One is if you are in the business of delivering wireless signals and you need to know where the obstructions are. In this case the deliverables will likely be a first return from the top of the obstruction and a last return from the ground.

The other specific case for first and last returns could be for transmission lines. However, in reality what most operators do for this type of survey is fly the LiDAR for the ground data, then narrow their beam down so they can get more hits on the wires and fly the line a second time for the wires. They typically configure their system to given them last return for the ground on the first flight and first return for the wires on the second flight.

But in 95% of cases the client’s deliverable is simply the last return. So it might be wise, if you don’t fall into those two above categories, to simplify things by saying you want a DTM delivered of the ground, and let the LiDAR operator do the rest for you. Typically all LiDAR systems are developed with software which filters irrelevant data.

Now, as a client, I also want to be sure that the company I contract is not going to fake the data. So I will want them to tell me what equipment they are using, how it has been calibrated and how they will use it. I will probably want to know how large their spot is when it hits the ground. I will also want to know what their quality control procedures are.

In the meantime, I will probably want to let them know if I contract them to deliver data accurate to one foot RMSE vertical and two feet RMSE horizontally, then by goodness, I wont pay them unless I get it!

So your specifications might be simplified considerably by stating in your own words, something to the effect that you want:

A company using a LiDAR made by a proven manufacturer

A last return only (unless there is a very compelling reason to state otherwise).

An accuracy of some specific precision horizontally and vertically in open hard-surfaced areas (not forgetting that no LiDAR systems can currently guarantee better than 6 inches absolute accuracy in any dimension).

Proper ground control to support the above accuracy.

And data deliverables in terms of a specific quantity: all collected data, data at a grid of ten feet, twenty feet or whatever (also not forgetting that in heavily wooded areas or in cities, a lot of the ground is covered and you will not necessarily get a ground point each time the laser fires).

Then, in addition, you probably should ask for an explanation of the methodology, the contractor’s method of quality control and ground truthing, its experience (how many years it has been doing this stuff?), its staff (how long have they been doing this?), both ideally longer than one year, and some references.

Of some concern may be the contractor’s ability to stay on the job, especially if it’s a large project. However, any company who has laid out over $1M to buy a LiDAR system, probably has something going for it (although it could be heavily in debt!). As a footnote to that, though, because these are expensive systems to operate and maintain, it is reasonable to expect to pay some money in progressive milestone payments for the service.

If there are other components to the project and LiDAR is only a part of the project, don’t be surprised if your preferred consultant has a LiDAR subcontractor. Not many companies can afford the outlay for a bit of equipment that isn’t going to work full time (10 or 11 months a year) for them. There’s no shame in that. After all, most of you don’t personally have a lawyer on retainer; you hire one the couple of times during your life when you need one.

Those of you who have read this far will have realized by now that the specifications I have suggested are tied to a deliverable – not to how it is actually acquired. A lot of people may think I am being too simplistic and they may be right, but as a customer aren’t those the real criteria of interest?

This is probably going to leave me open to haranguing by the techies in the industry. In response, I can say I believe there will be definite advantages to having guidelines for LiDAR operators, as these are in the interests of the customer - and these will come (from people like the ASPRS). But in the meantime, if the client gets the product he asks for, and he goes out and spot checks it in open, hard surface areas (let’s be fair here) to ensure it is really what it is supposed to be, then everyone will be satisfied and we can all go home happy.

First published in Earth Observation Magazine (EOM) April 2001