• CUWB 3.0 (Bernoulli)
Anchor Positioning and Survey

CUWB Anchor Deployment: Measure Twice

Introduction

This App Note touches on a couple of subjects for a successful anchor deployment. These are Planning, Positioning and Survey. The Survey Template may be useful for keeping track of everything.

Warning: When testing a new system the anchors may be moved to different positions before the final layout is achieved. We suggest that you temporarily mount each anchor until it has been tested in its assigned location.

When comissioning a system the general steps are:

  • Make a plan, unpack your anchors ard record their Serial Numbers.
  • Mount and power some anchors.
  • Conduct a quick survey of these units and enter them into the CUWB manager.
  • Test anchor communications and tag visibility with the CUWB viewer.
  • Adjust the placement of your anchors and fill in the density of the deployment as needed.
  • Finish the survey of all anchors. Update the CUWB manager, and set the tracked volume bounds.

Anchor Planning

Anchor planning involves the particular volume you’re working with and its available lines of sight.

  • Start with an understanding of your tracking area. A scale drawing of the floorplan is most useful for this exercise.
  • Put dots on the drawing where you will want anchors. Start with the periphery of the volume, and place place units on 10 meter spacing (or less).
  • Fill in the middle of the tracked volume (still using a 10 meter grid) Higher or lower anchor density may be used depending on the physical ostructions in the space and subsequent lines of sight. Try to keep 4 or more anchors in view of each tag position.
  • Where you need the best z-axis resolution make sure that there is an anchor directly overhead.
  • Give some consideration to how the anchors will be connected.

Anchor Cabling

Anchors participate in system measurements, and are also responsible for consolidating and communicating the captured data to the calculation server. Subsequently anchors are always running and require uninterrupted power. Communications and power distribution to anchors is a defining characteristic of any UWB system. Presently all CUWB anchors use Ethernet for their data and power connectivity. The anchor modeles being used will determine the cable installation that is required to support them. DWETH111 (POE) anchors will each require their own active POE powered drop. DWETH101 (CPE) anchors can be daisy chained in runs of 6-8 units, and powerd from a single passive POE Injector. Both DWETH models are approved for indoor use only. If elemental exposure is expected they should be covered or have additional protection applied.

Anchor Positioning

Anchor positioning can help mitigate or avoid signaling errors by establishing and maintaing good measurement diversity within the system. Radio based measurements must handle errors from multiple sources. Some of these are: Line of Sight Blockage, Multipath Reflections, Signal Phase Patterning, and RF Communications issues. Measurement diversity overcomes these errors by taking multiple measurements from multiple places, it is a key technique to achieving repeatable results with radio based technologies. When positioning your anchors try to create measurement diversity by observing the following guidelines:

Increase the number of anchors used: Math requires that at least four anchors be used to create soloutions in XYZ space. The useful number is really 8-10 for redundant deployments in an open spaces.

Spread the anchors out, and vary their deployed heights: The goal here is to create a diversity of angles between the fixed anchors and mobile tags. Distribute anchors outward around the space that you want to work in and vary their height to improve the measurement repaetability and Z axis arrulary.

Take a lot of measurements: This takes advantage of the and high performance characteristics of the CUWB system. By increasing the measurement rate or adding anchors, more data becomes available to the server, and it can subsequently generate improved results. Positional averaging can also be turned on to smooth the results even further.

Elevate anchors as much as possible: A “Line of Sight” is an imaginary line between a tag being tracked, and an anchor that it communicates with. It is also the systems measurement line of interest. Radio signals exchanged along these lines do not penetrate people very well, or metal objects at all, so elevating anchors to a point where the Lines of Sight are not being blocked is important for completing these measurements.

Mount anchors at a distance from their mounting surfaces: Metal in the form of drop ceiling grid, metal wall studs, or rafters, can cause periodic reflections that interfere with an anchors radio performance. In these cases, Ciholas recommends that DWETHS be mounted 100-150mm off of the offending surface. Suitable DWETH mounts are available through the Ciholas WebStore.

Mount surface common anchors at similar heights: DWETH anchors don’t “see” below their horizon as well as above, so to maintain communications any anchors that are mounted on a common surface should be height matched to each other.

Orient antennas towards the volume of interest: Consider that different Antennas can have different signal patterns. DWETHs for instance radiate from the top dome, but not through their bottom. So DWETH units near the ceiling should be mouned with their domes facing downward. Similarly if mounted near the floor their dome should be up.

A word about Master Nodes: Bernoulli systems work on a distributed timing scheme and no longer require a Master Node. Prior (Archimedes) systems do require a Master Node for central timing and configuration distribution. If your plan on deploying this type of system, you will need to add one more unit to your anchor budget. This Master Node needs to have good radio visibility to all other nodes and should be in a physically stable and centralized location. It usually works best to elevate the Master Node to improve its line of sight and to place it a little lower than the other anchors to ensure that everyone can communicate with it.

Perform an accurate survey: Inaccurate surveys can introduce offsets and jitter into the final position solutions. This can happen as the solution engine tries to integrate data from both “good” and “bad” anchors positions. As anchors are occluded or “leave the fight” a tags position solution might temporarily improve, but will jostle around again once the offending anchor is back in view of the tag.

Antenna (Anchor) survey

A good survey accurately captures the installed antenna positions and ultimately determines the quality of the systems location output. The X,Y, & Z positions of every anchor antenna are measured and entered in the CUWB manager. The Measurements are most easily taken with a Total Station, but effective measurement can also be made with a tape measure, and plumb bob, it just takes more time. The following pointers can help you conduct a good survey.

Record the unit serial numbers as you unbox them: An accurate survey begins as you unpack your hardware and record each of the unit serial numbers. The unit serial numbers can be recoverd after the anchors are installed; See Serial Flash, but it is much simpler to record them at the beginning.

Make a drawing of where each unit is: Along with capturing the anchors relative locations, a drawing can be beneficial in diagnosing signal coverage or cabling issues. A top down view is usually easiest, and it is best if the dimension scales are consistent.

Select an origin point and mark it on the drawing: The origin (0,0,0) will be the point of reference for all subsequent measurements. Ideally it will be well marked, and immovable in the “Real” world. The interior or exterrior corner of a set of walls makes a nice reference.

Measure to the center of each antenna: Relative to the origin that you selected measure to the nearest millimeter where the center of each anchor antenna is located. For DWETHs the functional antenna center is inside of the center of the dome. Usually measurements are taken from the origin to the middle surface of each anchor dome, then they are adjusted to be 24mm inboard from that point. On the drawing, record the (XX.XXX, YY.YYY, ZZ.ZZZ meters) positions to three decimal places. Measurements can be accomplished with a plumb bob and tape measure but a laser range finder or theodolite (Total Station) is usually quicker. The most important thing is to be consistent on how the positions are measured.

Initialize the CUWB Server: Measurements are only good once they are applied. Take the following steps from within the CUWB manager.

  • Working from your survey, enter the previously recorded anchor antenna positions.
  • Set the bounding box limits, so that the maximum reported tag positions are within the surveyed volume.
  • Z-axis calculation improvements may be realized by keeping the maximum tag extents to 0.5 Meters (or more) below the highest anchors as this preservs the Zaxis angular diversity.