More IMU Details
The list below is a brief breakdown of IMU's available for ScanLook users. Not all are included in our pricing page because we don't think they are absolutely necessary. The list is sorted by vendor.
| Unit | Maker | Hz | GPS | Acc1 | R/P2 | Hdg3 | Type | Weight(g) | Size(mm) | Bias | ARW | Notes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| xNav250 | OxTS | 100 | L1/L2 | 2 | 0.050 | 0.150 | MEMS | 425 | 132x77x41 | 5.00 | 0.200 | Self-logging, 4GB, < 10W |
| xNav200 | OxTS | 100 | L1 | 50 | 0.050 | 0.150 | MEMS | 365 | 132x77x36 | 5.00 | 0.200 | Self-logging, 4GB, < 10W |
| STIM300 | NovAtel | 125 | L1/L2 | 2 | 0.015 | 0.080 | MEMS | 500 | 152x137x51 | 0.50 | 0.150 | <4.6W |
| FSAS | NovAtel | 200 | L1/L2 | 2 | 0.008 | 0.012 | FOG | 2100 | 128x128x104 | 0.75 | 0.100 | 16W |
| ADIS | NovAtel | 200 | L1/L2 | 2 | 0.035 | 0.150 | MEMS | 475 | 152x137x51 | 6.00 | 0.300 | 2.5W |
| ISA–100C | NovAtel | 200 | L1/L2 | 2 | 0.007 | 0.010 | FOG | 5000 | 180x150x137 | 0.50 | 0.012 | 18W |
| ATLANS-C | IXBlue | 200 | L1/L2 | 2 | 0.005 | 0.020 | FOG | 2600 | 160x160x113 | 0.10 | 0.005 | < 22W |
| VN-200 | VectorNav | 800 | L1 | 250 | 0.100 | 0.300 | MEMS | 16 | 36x33x9.5 | 10.00 | 500mW | |
| MTi-G-700 | XSENS | 2000 | L1 | 200 | 0.200 | 1.000 | MEMS | 55 | 57x42x23 | 0.2 | 10.00 | 675-950mW |
1 Accuracy in cm
2 Roll/Pitch in degrees
3 Heading in degrees
There are three basic technologies used for IMU's: MEMS, FOG, and RLG.
MEMS, Micro Electro-Mechanical Systems, uses silicon chips with integrated tuning forks as sensors. These are the most popular sensor of the three, are the least expensive, and also the least accurate. They are typically small, rugged, and consume little power. They are generally inexpensive to produce, have a great future potential, are reliable and have a low maintenance schedule, and have a tremendous potential for miniaturized solutions (UAS).
FOG, fiber optic gyros, consist of a coil of optical fiber using the interference of light to measure angular velocity. They are better than MEMS and have reached and even surpassed RLG sensors performance. They have a very long life relative to other technologies. They are typically more expensive than MEMS and not as expensive as RLG sensors.
RLG, ring laser gyros, were generally considered to be the most accurate, but this changes over time of course. They are generally quite expensive.
RLG and FOG sensors are both based upon the same basic theory. The essential difference is that the RLG laser beams do not use fiber optics but instead use mirrors to direct the laser beams. In both sensors, the larger the better for accuracy as the path of light is increased. Both start up very fast as they are solid state devices (no moving parts).
As the excellent articles below show, while RLG-based sensors have established an earlier and larger market footprint, FOG-based sensors evolution has continued. Today FOG reaches its “ultimate theoretical performance and surpasses its well-established competitor, the ring laser gyroscope (RLG).”
http://www.ixblue.com/m/publication/ultimate-performance-fog.pdf
http://www.ixblue.com/m/publication/fog-key-advantages.pdf
More good IMU reading:
Emerging MEMS IMU and Its Impact on Mapping Applications by Naser El-Sheimy, Calgary
What is INS?
Basic Principles of Inertial Navigation
http://www.insidegnss.com/node/3123
http://www.insidegnss.com/auto/julyaug12-Goodall.pdf
http://www.insidegnss.com/auto/IG0307el-sheimyFinal.pdf
Outages and Accuracy
So now you know what we offer and how INS solutions work, now is a good time to learn just a bit more.
There are two key things we need from our INS. First, we need good angular accuracy. Second, we need it to help carry us through GPS outages. We absolutely need it for both of these until some other solution presents itself. Looking at the angular accuracy we can reverse out the needs of our IMU by the accuracy required of our final product. For a moment, let's say we need 1cm accuracy at 15m of range. Using the arc tan (1 / 1500) we get 0.04 degrees. You could then argue that we need 50% of that, or maybe 33% of that, but generally let's say that we need 0.04 degrees of accuracy in the tip/tilt (not heading). Ok, that is relatively easy. Now we can shop for an IMU that meets our needs.
Next up is the problem with outages. This one is not nearly as easy to determine and in fact you are left to the wiles of the vendor to specify what happens during an outage. Of course, you could purchase a unit and perform real world tests of your own but if you have enough money for that you should be on a beach somewhere relaxing or working at a home for the needy, but definitely not digging into high tech electronics trying to make a living. Let's take a peek at what our NovAtel information reveals. We're going to assume a post-processed solution using Inertial Explorer (IE).
| Outage Duration | Position Accuracy (m) RMS | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| ADIS | ISA–100C | FSAS | ATLANS-C1 | STIM300 | ||||||
| Horizontal | Vertical | Horizontal | Vertical | Horizontal | Vertical | Horizontal | Vertical | Horizontal | Vertical | |
| 0s | 0.010 | 0.020 | 0.010 | 0.015 | 0.010 | 0.015 | 0.01 | 0.01 | 0.01 | 0.02 |
| 10s | 0.040 | 0.020 | 0.010 | 0.020 | 0.020 | 0.020 | 0.02 | 0.02 | 0.02 | 0.02 |
| 60s | ---- | ---- | 0.040 | 0.030 | 0.130 | 0.050 | 0.15 | 0.01 | 0.23 | 0.03 |
1 According to IXBlue, ATLANS-C GPS outage data is based on demonstrated field measurement. During same measurements, ATLANS-C proved to be superior against other solutions for GPS outage test, despite stated theoretical spec of other solutions.
From the table above we see that all five units perform essentially the same if there is no GPS outage.
