Fullscreen HTML5Med Res (??MB)HTML4

A year on from the 2009 Victorian Bushfires, a possible life-saver goes begging . . . Defined Area Early Warning System You can sense the frustration in this first-hand account of the design and development of an effective, targeted and most of all affordable natural disaster warning system. Developed specifically in response to those terrible bushfires of a year ago, everyone who sees it says “fantastic!” So why is it languishing in some bureaucrat’s “in” tray? I t is difficult to travel anywhere in Victoria without coming across reminders of the fires that took so many lives on Black Saturday, February 7th 2009. The tragedy in one way or another touched us all and the depth of the generosity shown to those directly affected was truly admirable. Like so many others I also wanted to do something that would help prevent loss of life when, as it surely will, similar circumstances arise in the future. As an electronics engineer I thought that there must be a better way to get information to those in areas of high risk, to help them make the right and timely decision about when to activate their prepared fire safety plan – that is, whether to leave while it is still safe to do so, or to stay and defend. The inspiration. The inspiration that I was hoping 22  Silicon Chip would come finally arrived on 14th October, when I heard on the news that as part of the Unified National Bushfire Strategy there would be firstly; a common Fire Danger Rating (FDR) and that secondly; the ABC was to be responsible for broadcasting the FDR and fire threat warnings by radio. I immediately thought of using radio to carry the warning information in a digital format to a dedicated receiver to those in areas specifically at risk of fire at the time. As any designer knows, the inspiration is the easy part... current threat level for the area that it is physically located in, • Be inside the radio footprint of a carrier that would remain reliable throughout the duration of the emergency, • Have backup power to continue to provide service in the case of loss of mains power, • Be operationally reliable, physically robust and suitable for low-maintenance installations in rural areas, • Be easy to manufacture, cheap to buy and easy to install. What would it look like? The choice of carrier The initial concept of the receiver included the following requirements: • Be able to display in real-time the I had heard something of Telstra’s telephone-based National Emergency Warning System (NEWS) system by this time but thought that it had some severe limitations that would be hard to overcome. It seemed to be intrinsically complex and I felt that local infrastructure damage and high traffic use by David Ambry Engineering Manager, Nexus Technologies Pty Ltd siliconchip.com.au The Sam receiver, developed by David Ambry and the team at Nexus Technologies, as part of the DAEW System, can be used as a stand-alone installation or may be supplemented with external sirens and visual threat-level indicators – or even control automatic spray and sprinkler systems etc. in emergency situations would make it particularly vulnerable to failure, just when it is most important that the messages be delivered. I felt that augmenting the system with another delivery mechanism, something that worked in parallel with the Telstra system, would vastly improve the reliability of information delivery to those who required it, when they required it. My recent tenure designing satellite telephony systems based on Iridium resulted in this being initially considered a candidate but was quickly ruled out on grounds of cost. (It may well be being used by Telstra to deliver SMS where there are currently no copper connections.) Next up for consideration was the broadcast band AM radio, as it is ubiquitous in rural Victoria and by nature of the long wavelengths used, propagates well. A bit of investigation revealed that there was an augmentation of AM that provided a data channel – Amplitude Modulation Signalling System or AMSS. Unfortunately AMSS has not been siliconchip.com.au embraced by Australian broadcasters and also suffers from a very slow data rate. Therefore it was ruled out as a contender. The newly introduced Digital Audio Broadcasting (DAB+) service was also a candidate, albeit only briefly. As DAB is intrinsically digital, providing a data channel for threat messages is easy but there are other limitations that ruled DAB out. For example, the high frequencies involved with the DAB band (174 to 239MHz) make it affected by diffraction, absorption and multi-path effects, meaning that reliable operation could not be guaranteed for anything less than a line-of-sight installation between transmitter and receiver. But most importantly, DAB is currently only available in urban areas – no use at all to the high-risk areas that need to be served. That left the FM band as the one serious candidate. The FM standard includes RDS or Radio Data System, a method of transmitting data by way of a separate sub-carrier that is a simple enhancement of station equipment and importantly, has been widely adopted. An analysis of the available data on transmitter coverage and fire risk estimate maps showed that an estimated 90% of the population in the higher fire risk areas would be capable of receiving ABC FM transmissions. The rest was simple. The centralised co-ordinated controlling bodies would provide the fire threat information for the state, the FDR information delivered to the transmitters by Internet and then broadcast to the special receivers that would display the information. There was nothing technically that could not be done. The State Monitoring Service receives the locality-specific fire information and assigns a threat level to the relevant cells. The information is then delivered electronically to the FM transmitters and transmitted to the Sam receiver. March 2010  23 The Prometheus demonstration application runs on a PC and connects to the FM transmitter by internet. Each and every cell is able to be assigned its own threat level. The soft sell With the complete design mapped out in my head, it was time for the next hurdle – trying to sell the idea to my employer. Explaining the system and the building block elements that we already had to the owner of the company went surprisingly well. He was very receptive to my idea and I was allowed time and budget to fast-track a trial system. I am a senior Design Engineer at Nexus Technologies Pty Ltd, an Australian company that designs and builds high-quality, low volume Audio/Video equipment. One of the products that we already manufacture and sell is an FM tuner with RDS that I designed several years ago. A problem remained however: how to get the FDR messages only to those in areas that were under threat of bushfire attack. The initial proposal was that only those in the listening area, the radio ‘footprint’ of the given transmitter would receive the threat messages and those outside this area would simply not receive the data. Adjacent area transmitters would use different frequencies and receivers would be programmed to only listen to their stations. It would work but the coverage areas would be poorly defined. A method that provided a welldefined cellular division of area was 24  Silicon Chip required. I decided to break the state up into an x-y grid with each area thus defined being given a unique designator address. This also allowed for variable-sized cells; there could be smaller cells in high populationdensity, high-risk areas and larger cells in low-density, low-risk areas. Each receiver would be allocated its cell designator at installation time as well as the frequency that it would be tuned to. Indeed, the system now became frequency-independent so that the receiver could have a list of backup frequencies to attempt in the case that the primary radio station should fail. The transmitters would continually transmit each and every cell’s designator and associated FDR threat level. The receiver would decode all messages but only display the information for its programmed area. The perspiration The design concepts and outline were well received by the rest of the engineering team and we began to prototype the system. The project was given a suitable acronym – the Defined Area Early Warning (DAEW) system and consisted of the receiver (codenamed Sam) and the software to drive the transmitter (codenamed Prometheus). Aran Gallagher, a talented hardware and software engineer, rapidly developed the Prometheus software while Eddie South, indispensable engineering technician, sourced a low-power RDS-capable FM transmitter (commonly called an exciter) and set up an aerial, counterpoise, RF attenuator and Ethernet to serial bridge. I adapted our existing tuner design to deliver the RDS messages via serial port to an external display and decoder and within a month of elapsed time we had a crude but demonstrable system. Initially the system only displayed the standardised FDR text for each fire threat level on a single-line 16-character Liquid Crystal Display. It’s a poor design that doesn’t achieve any enhancements during the normal course of development and naturally the DAEW improved during the process resulting in a second-generation or pre-production Sam receiver by late November 2009. The improvements included the following: • The addition of an audio amplifier and speaker that would be automatically switched on whenever the FDR escalated to deliver any audible messages that might accompany the change. • Front-panel high-intensity 10mm LEDs in the appropriate FDR colours that would light to show explicitly the siliconchip.com.au FDR for the receiver’s cell. • Amother LED was also added to indicate a day of Total Fire Ban for the cell. • Drivers and connectors for an external siren and status display. • Finally, uncommitted relay contacts were included. The relays are energised when the threat level is above the Low/Moderate level and at the Catastrophic level, the idea being to drive auxiliary equipment such as a roof sprinkler system. If transmitted, each cell can have its Time To Impact (TTI) estimate transmitted, though this doesn’t need to be done as often as the FDR data as the receiver can run an autonomous countdown timer to keep the display of this estimate updated. We also concluded that cells could be updated at different rates and that we could be transmitting the FDR information to critical cells more often than to those that were under no threat. Sam uses a flexible software guard timer, with different decision thresholds for different FDRs, to advise of any loss of received code in the allocated time. This naturally led us to propose self-escalation of FDR in the event of signal failure but this was deemed unwise for obvious reasons. As the receiver recovers each and every cell’s status information, we also decided that the complete state situation map could be recovered at any receiver site using software that is something like the inverse of Prometheus connected to the serial port of Sam. This may be of use to local authorities that need to be advised of the overall situation. The Sam receiver is housed in a simple sheet-metal enclosure, being robust and cheap to fabricate. It might not win any styling awards but this is a case of form simply following function. The overall dimensions of the enclosure were driven by the need for a large internal speaker for good audio efficiency and the size of the internal SLA backup battery. The power input accepts any voltage from 8 to 40VDC to charge the SLA battery, so will accept power from a 12V plug pack or 12V to 24VDC Remote Area Power Supply (RAPS) source. A 1m fibreglass whip antenna is used as the receiver’s aerial – plugging straight into the antenna socket, although a remotely-located high-gain siliconchip.com.au Yagi might be required in areas where the RDS signal is marginal. We had trimmed the core design of the Sam receiver down to a level where we had a viable business model that would deliver a small but acceptable profit margin selling the basic unit at a price of $100 on a production run of no less than 10,000 units. That was the easy part! The design complete and tested, suitable for demonstration and the value of the solution undeniable, we thought we had a system that could not fail to generate widespread interest. We had ready access to CFA volunteers and demonstrated the DAEW system to them with encouragingly enthusiastic responses. It was time to see what the government and the appropriate authorities thought of it, introduce it to the media and see if the ABC would embrace the concept. We enlisted the help of Louis Delacretaz, former Mayor of the Shire of the Yarra Ranges, a qualified engineer himself and someone extremely concerned about raising the fire threat awareness level of people that live in these areas. Louis immediately understood and embraced the DAEW system concept and agreed to help champion our cause with the authorities. A meeting was held in early December with James Merlino MP at Parliament House. James lives in Monbulk, right in the middle of the Dandenong Ranges high fire-risk area and was Louis’ choice as he felt James might best understand how the system would serve the community. The DAEW system was favourably received at the meeting and we were asked to provide a detailed system description and report which would be presented to the Premier of Victoria, John Brumby. We are now waiting on a formal response to our system implementation proposal. We have made tentative steps in making the media aware of the DAEW system but have yet to generate any interest. It would seem that just having a better mousetrap doesn’t mean that people are automatically going to be interested in it. We also approached the ABC to determine their capability to provide the RDS transport mechanism. Submitting a formal query resulted in a wait of almost a month before we were provided a response: “In short, no, the ABC doesn’t support RDS for any of our radio broadcasts in Australia... the ABC experimented with the technology in the early 1990s but back then, not many receivers were able to decode the broadcasts.” We simply cannot believe that the ABC has exciters that do not have RDS capabilities. The cost of RDS subcarrier functionality is insignificant compared to the rest of the infrastructure costs of FM transmission. If their equipment is more than 30 years old, it may not have this function but I’d be prepared to bet that all their station racks have an exciter with a DB9 RDS input socket that is connected to nothing. Where to from here? We were, in all honesty, naively optimistic that the DAEW system would be rapidly and enthusiastically embraced. After the Internet information site fell over at the first whiff of summer smoke we were seriously considering whether we could roll out perhaps a thousand units by February for immediate deployment in some of the highest risk areas. If the ABC couldn’t or wouldn’t assist, maybe regional and community FM stations might be willing to host the RDS traffic. Compared to last year it’s been a cooler and wetter summer but it’s always late summer; February and even into March that are the driest months and present the greatest hazards. Here we are – at the time of writing (end January) and we’ve not had a word of response from the government. By the rules of simple good manners, that would indicate that DAEW is still being given serious consideration. We can but live in hope! Further reading w w w. c aw c r. g ov. a u / b m rc / w e fo r /projects/fire_wx_workshop_ jun_05/08gould.pdf (The map on page 3 shows the potential for disastrous bushfires). www.cfa.vic.gov.au/about/documents/7th_Feb_2009_Fires_Overview_Map_A3 (An Interim Summary Overview map of the 7th February 2009 fires shows the predictions to be very accurate). SC March 2010  25