Development and installation fo a WSN physical security system

Summary

This deliverable presents the work carried out to date on the design, installation and test of a Bluetooth, proximity-based RTLS (Real Time Location System) for use as an ICT physical security layer for use in Boston Scientific’s Clonmel facility. Prior Art research had been carried out for thisusecasein D4.6Viability of WSN as ICT Overlay for Physical Security Detectionof which this document is simply a continuation with little cross referencing and so is not covered here. All work has been carried out by Tyndall in collaboration with Boston Scientific (BSL) who facilitated the test environment.There are a number of wireless tracking systems that can be used in an indoor environment. Previously D4.6 reviewed some of these and found that whilst highest accuracy is found in trilateration/triangulation-based systems, these systems are complex, relatively power hungry and expensive. In a large factory environment where tracking of materials and equipment is required over a very large area,a proximity-based solution is an attractive solution due to its low cost, low power characteristics. Large ware houses today typically use RFID as a proximity-based solution. Compared to using Bluetooth as a proximity solution these tend to be bespoke and higher cost.Every use case is different and so different technologies all have a place in solving these needs. In many cases it is a combination of technologies that provides the best solution. In the BSL case,tracking equipment and providing security for removed materials was suited to BLE. This report shows that the chosen off-the-shelf BLE solution selected the correct proximity (beacon) over 90% of the time. As one would expect with a proximity system,as the tag (the thing we are tracking) moves towards the boundary of two or more beacons at similardistance,the system performance degrades.This degradation is highly dependent on the physical environment as constructive and destructive addition of multipath signals can significantly boost the signal level of a more distantbeacon over that of the closest beacon. In one experiment it was shown that a region of about 90cm exists when 2 beacons are placed 4.2m apart on a planar surface where system performance is degraded. In addition,as this system relies on signal level to determine distance,the absolute output power of all beacons needs to be the same. Further to this,the radiated power is highly dependent on the antenna radiation pattern,and therefore the orientation of the beacon and tags can be a factor in accuracy.There are options to reduce the region of uncertainty such as the use of historical data,aggregation, inertial measurements, radiated power calibration and fingerprinting.This is all identified as further work.As we look to scalingup the system to thefull manufacturing floor,the maintenance of the system becomes significant. Managing battery replacement is a significant issue,and so this report includes work that has been done to reduce battery replacement in the beacons which are the hardest to reach devices as they are usually located on the ceiling. Currently the work shows a doubling of battery life using photovoltaic cells.Finally,during the course of the deployment work in BSL there has been significant interest in the technology and how it can help solve problems that exist there, showingthe power of providing a live demonstration in a real working environment