USE OF TETRA NETWORKS IN CRISIS SITUATIONS + LARGE SCALE EVENTS A study of communications in various crisis situations in Europe

www.kcglobaltetra.com KC Global TETRA Ltd 374, Coventry Road, Hinckley, LE10 0NH, UK

Phone: +44 1455 251348 Mob: +44 7765 660358 E-mail: kenn@kcglobaltetra.com

Chapter 1

Use of TETRA in Crisis Situations A study of communications in various crisis situations in Europe

BACKGROUND During 2009 the Netherlands public safety organizations encountered a number of crises where the efficacy of the C2000 system was called into question to some degree. As a result of discussions at the PSRG (Public Safety Radiocommunication Group: an informal group of European governmental representatives of national radio projects) it was decided to organize a meeting to discuss the use of TETRA in crisis situations with other European country’s public safety organizations, in order to establish whether any conclusions could be drawn as to how improvements could be made. The TETRA Association agreed to support this effort by asking Kenn Coates of KC Global TETRA Ltd to contribute.

NETHERLANDS CRISES The crisis situations arising in the Netherlands during 2009 were:

1. The crash of Turkish Airlines Boeing 737-800 on 25 February 2009 2. The attack on the Dutch Royal Family during the Queen’s Day Parade in Apeldoorn on 30

April 2009. 3. The riots at the beach party near Hoek van Holland, Rotterdam on 23 August 2009.

TURKISH AIRLINES CRASH The preliminary report by the Dutch Safety Board discloses that the aircraft crashed at 1026 hours on 25 February 2009, about 1.5 kilometres north of Schiphol Airport, Amsterdam, runway 18R as a result of the crew failing to notice that the autopilot system had severely reduced the speed of the craft on approach to landing.

Of the 7 crew and 128 passengers on board, 4 crew members and 5 passengers died and 3 crew members and 83 passengers were injured.

This was a spontaneous, unplanned, major incident.

ATTACK ON DUTCH ROYAL FAMILY At 1150 hours on 30 April 2009 an open top bus carrying several members of the Dutch Royal Family including Queen Beatrix and Prince Willem-Alexander, Prince of Orange and heir to the throne, was on a celebratory parade in Apeldoorn.

This was an annual event and the route was known and published in advance.

2

At the point where the bus made its last turn in the road towards Het Loo Palace is a road junction which was closed to other traffic using light metal barriers; a Suzuki Swift motor car driven by Karst Tates, a 38 year old male Dutch national, drove through the barriers at speed – apparently in an attempt to collide with the bus.

In doing so the vehicle collided with a number of spectators and officials causing the death of 8 people and leaving 10 others injured.

The driver of the car then lost control of the vehicle due to his own injuries and it came to rest after colliding with a Royal Monument, De Naald (The Needle).

Tates died of his injuries the following day.

The event was planned and there were a number of police officers in place to deal with traffic and crowd control, there were Royalty Protection officers surrounding the bus and ambulance and medical staff on hand to deal with any medical emergencies that might arise where large numbers of spectators are gathered.

The attack on the Royal Family could not have been anticipated and this turned the event from a planned event into a spontaneous major incident.

RIOTS AT BEACH PARTY, HOEK VAN HOLLAND On 23 August 2009 a planned music and dance festival – Veronica Sunset Grooves - took place on a beach near the Hook of Holland.

It appears that permits had been issued to allow 15,000 people to attend the festival, which was a regular annual event, but that nearer 50,000 people had attended making crowd control difficult.

There were 45 police officers on duty and, despite the fact that intelligence had been received prior to the event that football hooligans intended to attend to make trouble, criticisms have been made in the Dutch news media that the local police chief had refused to authorize the attendance of specialist riot police in an attempt to reduce costs.

It is understood that fighting broke out about midnight and officers at the scene summoned assistance, but the roads approaching the area were choked with traffic – making it difficult for reinforcements to reach the scene; a group of officers, under attack and in fear of their lives, drew their firearms and opened fire on their attackers.

One man was killed and six others injured by the gunfire.

Again this was a planned event, but the attack on the officers and their reaction turned it into a spontaneous major incident.

Chapter 2

ANALYSIS It is not uncommon, where there is an enquiry into the handling of a major event, operation or disaster situation, for communications to be criticised. A workshop was held in Brussels on 23 February 2010 at which, in addition to the incidents discussed above, a small number of presentations were given describing the experience of other countries’ TETRA systems when used in crisis situations. This was arranged in order to disseminate to the wider Public Safety community, using TETRA networks, the need for preparation in order to make best use of TETRA for communications in crisis situations. It was also thought that this might well lead to a need for further workshops concentrating on the constituent elements of TETRA, as well as related operational practises. It must be understood that in the instances focussed upon in this report, that is the incidents occurring in Netherlands in 2009, the author has not had full access to operational public safety staff and therefore some of the conclusions are based on the author’s own experience as a police officer in various UK organisations, his knowledge of TETRA and on news reports from news agencies based in the Netherlands and found on the internet – as well as the presentations given at a meeting in Brussels on 9 December 2009 as well as the workshop in Brussels on 23 February 2010. The Netherlands is organised in Safety Regions and has, or is in the process of moving to, tri-partite emergency services call handling and dispatching centres. There are 25 Safety Regions in the Netherlands, each with a Police Department and an Ambulance Department. The Fire & Rescue services are in the process of being reorganised along the same lines, but at the time of the Turkish Airlines crash there were 425 Fire Departments. Each Safety Region is run by Local Government, with a Mayor having overall responsibility but each of the Emergency Services having its own organisational structure. All of the emergency service agencies use the government owned and operated C2000 TETRA system for their radio communications and paging needs. The C2000 system has been in use since 2004.

TURKISH AIRLINES CRASH This incident took place in the safety region of Kennermerland, the second most critical region in the Netherlands with a population of 500,000 and covering among other things Amsterdam Schiphol airport, the Corus Steelworks and the sea port serving Amsterdam. The first calls regarding the air crash started to be received in the emergency services control room about 1030 hours on 25 Feb 2009.

2

The calls were being made by motorists who had witnessed the aeroplane going down, as well as from Air Traffic Control (ATC) who were reporting having lost sight of the craft on radar. The ATC did not know what had happened to the ‘plane, but motorists were reporting that a large aeroplane had crashed – though the location given was somewhat confused. The emergency services started to dispatch resources to the scene as soon as possible. Once it was confirmed that this was a full scale aircraft emergency resources were drawn from outside, as well as across the region. The author has been unable to ascertain precisely how many resources attended and from where, but it seems that a fleet of 60 ambulances attended from across a number of Safety Regions. The total number of public safety personnel at the scene was between 600 – 700. What is clear from investigations carried out by the C2000 network operation staff is that there were 8 base stations supporting the radio communications with an average of 3 carriers per site. The majority of traffic was borne by the Halfweg site, a 3 carrier base station. This site, at 1000 hours, was supporting 49 active talkgroups. By 1100 hours the same site was supporting 116 active talkgroups with a further 300 registering on it and this remained the position until between 1200 – 1300 hours. By 1400 hours it had reduced to 89 and by 1500 hours it was almost back to normal at 59 talkgroups. Radio users encountered congestion on the radio network, which they had never experienced before – as did dispatchers in the control room. This led the users to believe that there was a problem with the network. Because of this some communications moved from the radio system to the GSM system. It seems likely that this will have caused something of a breakdown in communications, especially as the C2000 system normally uses the All Start system setting in the fleetmap for its group calls. The network management centre was contacted about 1045 hours and placed an immediate block on maintenance operations. Extra personnel were also dispatched to the switch and the Network Management Centre as a precaution. Records show that the use of capacity on the Halfweg site shot up at the time of the incident and exceeded 90% until about 1230 hours, falling to 80 – 90% for about 30 minutes and then further reducing to between 70 – 80% for about 45 minutes. There was a further surge in capacity usage on this site to 90% about 1415 hours; it is thought this second peak was due to a visit to the scene by the Prime Minister, and/or moving injured people and relatives to a different location.

3

Following discussions between the Kennemerland control room staff and the network management team it was decided to surrender the reserved data slot on the system, freeing it to help cope with the demands for voice traffic. The system was also switched from All Start to Fast Start. Although it is likely this made the system more flexible, the fact that users were unused to it meant that there were some incidences of users asking for messages to be repeated where they had joined a call on Late Entry. It is significant that, other than the congestion reported above, the C2000 system did not encounter any other problems and carried on working throughout.

QUEEN’S DAY PARADE This incident took place in the town of Apeldoorn in the Gelderland province. It was mentioned in the presentation regarding this incident that it was unknown at the time whether this was an attack by an individual or part of a wider terrorist attack. The behaviour of the C2000 system has not been analysed in the same way as it was after the airline crash, basically because there have been no official complaints about the behaviour of the network. In the main, other than some slight congestion (less than 5%) problems, the system worked well; a small group of users – the Royal Family’s protection team - reported that they had encountered some strange system behaviour, but did not make a formal complaint. It is understood the problem encountered was that not all users in the group were receiving all of the messages that their colleagues did.

BEACH PARTY RIOTS This incident took place on a beach near the Hook of Holland. The party was a planned event, but there were disturbances after nightfall caused by football hooligans attacking the police. Police officers reported that they were unable to contact their control room on the C2000 system, and that their efforts to make emergency calls had been unsuccessful. They complained that the system had failed. The network coverage at the locale is known to be at or just above minimal, but the network did transmit 6 emergency calls and there was no apparent congestion – though a full investigation has not been undertaken to analyse the particular behaviour of individual terminals.

Chapter 3

INCIDENTS OUTSIDE NETHERLANDS In order to demonstrate that this was not an issue peculiar to the Netherlands, representatives from other countries were invited to present their experience.

DENMARK – UN FCCC COP15 This presentation showed the preparations made by Denmark in readiness for the hosting of the 15th Conference of the Parties to the UN Framework Convention on Climate Change in Copenhagen in December 2009. Copenhagen had hosted an IOC conference in October 2009 and found that capacity on the system was pressed to the limit. The Danish system is known as SINE and the police take a lead role in its running, being the contract holder and responsible for fleetmapping, capacity and implementation. As a result of the October conference, and in the knowledge that the December conference would also be held there, a working party comprising the police, Motorola and SINE was set up to prepare for the COP15. Additional capacity was installed at 7 sites, some had their capacity doubled. Denmark was fortunate in that Sweden had not yet fully rolled out its TETRA system, and Denmark was able to use some frequencies that will be used in Sweden eventually. The police moved into a call centre in order that they could act as a single point of contact for the other agencies using the network and the network operator. A meeting was held each morning during the conference where the behaviour of the network during the previous 24 hours was reviewed and advice was given to senior users. 90% of the police officers deployed at the December conference had not used a TETRA terminal before and were given less than an hour’s instruction. The system was not used as all informed, users could hear only messages from dispatchers – not messages from other users. A number of issues arose; often when a group of users were together in a vehicle not all of the terminals worked until the users dismounted from the vehicle and spread out. At large demonstrations when 4 or more large talkgroups were in use, communications broke down. One base station site switched to Security Class 2 – rather than Security Class 3 – meaning that terminals were unable to affiliate to the site. The ATIA log showed that a large number of terminals had been denied the right to affiliate to a site.

2

In certain areas of Copenhagen metallic speech was experienced. Only during the conference it was discovered that a large number of police officers were being accommodated on a ship in the North harbour, and extra capacity had to be installed at the supporting site overnight. Despite these problems the system appears to have held up well, though analysis is still ongoing.

TRAIN CRASH, HALLE, BELGIUM This incident occurred near Halle, Belgium on Monday 15 February 2010. Belgium uses the ASTRID (All Round (Semi) Cellular Trunked Radio with Integrated Dispatching) TETRA system. The fleetmap and radio configuration are designed in such a way that users are regularly required to change talkgroups when an incident occurs. The fleetmap has talkgroups for each area replicated across all radios, so that users know which talkgroup is to be used depending on where an incident occurs – no matter where they work. Belgium uses dispatchers on their primary talkgroups, but there are also unattended talkgroups that can be used if a dispatcher is not required. The overriding principle is – in an emergency do the same as in routine circumstances, i.e. don’t have special talkgroups for particular events necessarily – have talkgroups that are used for any special event, whether planned or not. This means that users know what to do in the first instance in an emergency. Once resources arrive at an incident – depending on the nature of the incident – they may be directed into a particular work stream and have to change to another talkgroup, but initially they will all change to the same talkgroup. Discussing the train crash it must be remembered that the workshop took place only one week after the event, but initially it appears that: The police reacted very quickly and switched to a provincial talkgroup as a matter of routine and attended what they call the First Destination Point (Rendezvous Point) – a meeting point where they would meet supervisors who would direct resources. The disaster plan was implemented quickly and the appropriate staff attended. Unfortunately the command staff are not as well prepared in terms of changing talkgroups as the field staff, no situation report was given to tell the control room that everything was in place. Also there is still a tendency among command staff to use GSM, rather than ASTRID. Multi-agency (or multi-discipline) talkgroups were taken into use quickly.

3

At the time of the workshop the use of the radio system by other agencies was still being evaluated, but it was said that while there was – technically – some queuing this was not really felt by resources as operational congestion.

FLOODING, CUMBRIA, UK NOVEMBER 2009 Following several days of heavy rain in the North West of England an area of Cumbria around Cockermouth had experienced severe flooding, resulting in a number of road bridges being damaged and – in one case – washed away. A police officer, PC Bill Barker, died when a bridge from which he was diverting traffic was washed away. The flooding resulted in damage to the areas infrastructure including homes, businesses, roads, power supplies and communications networks. Given that the UK Airwave TETRA network also relies on the power and Ground Based Telephone Network, the TETRA system used by the Public Safety agencies was also affected. Nevertheless it was found that because of diverse routing of the GBN and through the use of batteries and generators the TETRA network did carry on working. Unlike many major incidents, this carried on for a period of several days – rather than hours. This presentation also touched on preparations under way for the staging of the 2012 Olympics in London and other high profile events like Notting Hill Carnival and New Year’s Eve celebrations, policed by the Metropolitan Police Service.

Chapter 4

GENERAL Initial reports of major incidents are often reported as having occurred at a number of locations, due to their being reported by members of the public who do not know the area well and are shocked by what they have seen. It is inevitable that users involved in a major incident will experience stress and that this can inhibit their ability to perform unusual tasks. Users will feel a sense of urgency regarding their communications needs and will experience feelings of frustration if a system they can normally rely on appears not to perform efficiently at times of greatest need. They may feel a need to impart more information about what they are seeing or experiencing, leading them to speak more and contributing to the problems of others who are competing for air time. Younger users, who have grown up with mobile telephony as a common communications tool, seem less inclined to observe radio discipline. This tendency is also spreading to older users for the same reasons. It is almost inevitable that members of the press and members of the public will be attracted to the scene and may interfere with or contaminate what may well be the scene of a crime; even if it transpires that it is not a crime scene, it will be treated as a potential crime scene.

PROBLEMS ENCOUNTERED This report is concentrating primarily on the Turkish Airlines Crash, as did the workshop that prompted this report. This is because it is an excellent example of the type of major incident that can occur anywhere, at any time and exemplifies the type of incident that causes most of the typical problems encountered. Resources are drawn from over a wide area and arrive at the scene monitoring their ‘Home’ talkgroup. This causes a number of problems;

1. Large numbers of talkgroups register on the local base station. 2. Communications are fragmented because information is being delivered to several different

dispatchers, sometimes in different control rooms. 3. The same information is being delivered several times by different people on separate

talkgroups. 4. It is difficult to marshal resources, due to the breakdown of command and control and loss

of all informed communications. 5. The large number of talkgroups causes congestion on the system, creating a vicious circle

of users struggling to get through, multiple dispatchers receiving information and updating incident logs and multiple calls back out from the control rooms.

2

6. Users also used their GSM telephony to communicate, which worked quite well, but this creates a risk of loss of ‘all informed communications’ and, when using it for contacting the control room, using further resources in the control room to answer their calls.

In this case there was an emergency plan for dealing with an air crash, but it was predicated on the crash occurring at the airport rather than nearby. Although the plan had been tested, the testing appears to have been restricted to senior officers. The testing does not appear to have included testing of the communications system, though this has not been confirmed officially. The upshot is that the field based personnel were not familiar with the plan – again this has not been officially reported. As it appears that there was no effort to reduce the number of talkgroups in use at the scene it can be inferred that there is a lack of understanding amongst both field based and senior personnel as to how the TETRA system works. The users then assume that there is a problem with the network, and that it is not fit for purpose when it comes to dealing with a major incident. This sentiment is then picked up by the press, politicians and unions and adverse publicity reduces the faith of the users in the system.

Chapter 5

LESSONS LEARNED Looking at the 3 incidents in the Netherlands there are a number of lessons learned, some specific to one incident while others are common to all. Coverage – it is a common complaint of users that coverage is not universal, but this is inevitable given that there are a fixed amount of frequencies available for use in TETRA systems and that coverage requires bases stations – one of the most expensive components in the system. In the case of the Hook of Holland incident it is accepted that coverage is at or only slightly above a minimum level – though there are no reports of complaints regarding coverage during the day prior to the disturbances. The testing done – post incident – was carried out when the beach was empty. While it can be accepted in theory that the coverage itself will not alter whether the beach is empty or whether there are crowds of people present, it must affect the behaviour of radio terminals when there are 50,000 people occupying the space. It could have a similar effect as a wood in the winter compared with the summer when the effect on the radio signal is much bigger because of the grows of vegetation. There is no suggestion that lack of coverage was an issue in the other incidents. Capacity – the capacity available on each base station should be planned according to its anticipated usage including;

1. location (city centre, urban, suburban, rural), 2. special premises nearby (airport, football stadium, concert venue, shopping centre, etc.)

and 3. the number of resources routinely patrolling the area, the number of different talkgroups it

has to support and the pattern of calls (group, private, telephony). It would not be a sensible use of public money to simply say that every base station must have 4 carriers, just in case, no matter how lightly it is used in normal circumstances. The network operator can be expected to study patterns of usage and move capacity around after an initial period of usage has proved or disproved the accuracy of the planning stages. Capacity can also be temporarily increased to deal with an anticipated surge in capacity requirements for a planned infrequent event. However, at a spontaneous major incident – at least in the first few hours – users must deal with the system as it is, no matter how little capacity is available. It should be borne in mind that Direct Mode is also available if capacity is an issue; this could be used for groups concerned only with local duties, such as crowd or traffic control where contact with a control room is not essential.

2

Fleetmapping – there seems to be widespread confusion as to what fleetmapping is; to many people it is simply a list of talkgroups - and to the end user that may well be acceptable, as they only see it in their radios as just that. However, the process of fleetmapping is actually the study of an agency’s organisational shape and purpose, its command and control policies and procedures and the distribution of its resources and mapping or shaping the fleetmap so that it reflects the organisation’s methods of working and its communication needs in the set up of the radio system. During the Turkish Airlines incident it appears that someone, it is believed one of the system engineers, suggested switching the talkgroups from All Start to Fast Start. This was done in a few dedicated talkgroups (less than 5). These settings are one of a number of settings chosen for each talkgroup. All Start – this settings instructs the system not to allow a group call to go ahead – including a call being made by a dispatcher - unless each of the members of the talkgroup has capacity available on the base station on which they are registered. If there is capacity available on a site at which a member of the group is registered but not at others, then the system will reserve a traffic slot ready to allow the call to proceed when capacity is available on all of the sites. Fast Sart – this setting allows a call to proceed immediately, providing there is capacity available on the site at which the instigator of the call is registered. It is understandable that it might be thought advisable to use All Start to ensure that every member of a talkgroup can receive the whole of each and every message on their group. However, there are several arguments to set against this view;

1. In an area experiencing congestion All Start is likely to exacerbate the congestion problem. 2. A radio monitoring a talkgroup while situated in an area remote from the group’s

operational area could prevent calls from going ahead, even though the radio user is not available to respond to the message or – in the case of a vehicle set – the radio is not even being monitored.

3. Anyone with operational experience will know that users do not listen to every message on the group, filtering out calls that do not directly address or concern them, or do not sound urgent, or because they are busy speaking to someone else face to face, or they are concentrating on a task, etc.

Fast Start allows the system to work at its most flexible and Late Entry allows users who have missed the start of a call to hear the remainder of the message when there is capacity available. (Or if they have just come back into coverage or out of DMO, or have just turned on their radio.) Unfortunately where All Start is the norm, as in the Netherlands, when the system is suddenly switched over to Fast Start users hearing part of a message, when they are used to hearing all of every message, can contribute to congestion by requesting the repetition of the entire message; this occurred during the Turkish Airlines incident.

3

Control – this encompasses several elements. Control at the scene, primacy, as well as control of the radio network – which is what we are concerned with here. Once it became clear that there were congestion problems at the scene of the Turkish Airlines Crash someone needed to take control of the issue and determine what was causing the problem and work out what could be done to reduce the impact. As with most things, of course, prevention is better than cure. In Belgium the users, at least the police users – who are usually the most numerous, would – thanks to their training and the configuration of their radios – have changed talkgroup as soon as they were dispatched to what was clearly going to be a major incident. This would have automatically limited the number of talkgroups in use at the scene and gone a long way to preventing congestion. It would also have ensured all informed communications and allowed the control room to be fully informed. Operational – There are a number of sub-headings that fall under this area. Planning – there are several elements to this. The Belgian police model of users being aware that they need to change talkgroups at the outset of an unusual incident is the result of forward planning in both the fleetmapping and the radio configuration stages, and of the users being trained in what is expected of them. Also, as mentioned above, there was an operational plan created jointly by the safety region and the airport to deal with an aircraft crashing at the airport. The majority of air crashes do occur either at or shortly after takeoff and landing, so it is not unreasonable to plan for crashes to occur on the flight paths used. But no matter how good the plan it is unreasonable to expect field based staff to follow a plan that they aren’t familiar with. It is also strange that there seems to be a tendency to write emergency plans incorporating the use of the communications system but not to take into account that the communications network is not universally available to the same degree everywhere; as discussed earlier some areas have better coverage than others, while capacity also varies from base station to base station and together these 2 elements more than any other dictate the usability of the system and its ability to cope with the plan.

4

Training There seem to be wide variations in the time devoted to training users in the use of a new TETRA system, and what that training will encompass. In the majority of cases it seems to be training users in how to use a radio terminal; but TETRA is so very different in its behaviour to an analogue system that it seems perverse not to teach users something of how TETRA works and what potential problems may be encountered. Users are trained only in the everyday use of the system, not what they should do in the case of an unusual or major incident. In the case of the Turkish Airlines crash it is said that the users had no experience of queuing or congestion on the system, but this is one of the main benefits of TETRA – that it doesn’t depend on being the first to get in when there is a gap in traffic, the system itself orders the queuing according to priorities set in the fleetmap. Many users just aren’t aware that you should press and hold the PTT; they release and retry, and this again contributes to congestion as the system is having to deal with several call setup attempts from each user. Some users seem to think that if they can’t get through – even if it is because they are without coverage – pressing the emergency button will somehow connect them to the control room. This may well have been a factor in the Hook of Holland incident. Also, regarding the emergency button, the ‘norm’ seems to be that the emergency button has to be held for 2 seconds before an emergency call is activated. This has been done to avoid inadvertent or accidental emergency calls being made, which is understandable; where a 1 second press has been tried it has been found that the emergency call is accidentally triggered too frequently. But 2 seconds sitting at a desk and 2 seconds when one is in fear of one’s life, or urgently trying to summon assistance for a member of the public can seem like very different periods of time. There is also the fact that radio discipline seems to be disappearing as a concept. The tenets of Accuracy, Brevity, Clarity, Rhythm, Speed, Volume and Pitch aren’t taught anymore; consequently there is a growing tendency toward rambling speech even on group calls, and this also contributes to problems with capacity and costs of operation as messages take longer than required to pass. Users should also be taught to use radio procedure even in a private call, as this is still a radio call – even though it has some of the characteristics of a telephone call. The biggest training problem though is probably that training is seen as a one-off, with training given at the outset of a TETRA system being taken into use and never repeated unless there is a change in terminals – if then. Users will be regularly trained and tested in the use of firearms, riot control, changes in legislation, fire fighting and medical treatment appropriate to their profession – yet the one piece of equipment which they may depend on more often than any other – their radio – is treated as an afterthought until it fails to perform as expected. Perhaps it is time to adopt a testing regime whereby a minimum standard of proficiency must be achieved in order to qualify as an operational user.

5

Exercises Where plans are tested, there seems to be a tendency towards carrying out table top exercises, where senior officers are given a set of circumstances and asked to say how they would react to various turns of event. This again does nothing to prepare the field based staff or to test the ability of the communications system to cope under stress. Interoperability/Multi-agency Working The idea of all, or a number of, agencies sharing the same communications network is a sound one in terms of financial and planning efficiency. Yet the idea that different agencies, performing often very different types of duty, will use the network differently appears not always to be given any thought. At the Turkish Airlines crash scene there were in the order of 60 ambulances, each containing at least one radio terminal monitoring its own talkgroup. But a talkgroup is a virtual radio channel and is intended to allow groups of people to communicate because they have a shared organisational aim, or they are responsible for a geographic area or a combination of those. While it is understood that this model is used to preserve patient confidentiality in normal circumstances, it appears not to have occurred that the model would impose stress on the communications network in a major incident scenario where large numbers of ambulances are gathered close together. The use of one talkgroup for initial contact and private call for discussion of anything confidential would be one model that could be considered. It was said after 9/11 that the emergency services do not have a culture of speaking to one another by radio, and there can be no doubt that this has grown up around the fact that historically each agency would have its own separate network; yet the advent of shared networks seems not to have encouraged the thought that this allows the agencies to work differently – if there is an operational need. If a joint plan exists for the response to an incident then it needs to be jointly conceived, written and tested – and if this is a new model of working then it probably needs to be practised regularly to ensure that each agency is aware of each of the others’ working practises and procedures to ensure they are not going to conflict with one another in reality.

Chapter 6

POSSIBLE SOLUTIONS Possible solutions to the problems encountered fall into a number of areas, as with the lessons learned above. Coverage In the main coverage is planned at the project stage when introducing the TETRA system and will seek to overcome holes in coverage in the existing system. Coverage requirements need to be specific and not percentage based, and they ideally need to be demonstrably testable. Where there is a spontaneous incident the coverage has to be accepted as it falls, but the use of Direct Mode together with gateways and repeaters may help. But where there is a planned event or an area that is more at risk of incident – like an airport, or a petro-chemical works, football stadia, etc. then the coverage should be planned to give good overlapping coverage if possible. Where there are events planned in unusual venues, like the Hook of Holland music festival, it makes sense to test the coverage there beforehand – if possible before permission is given for the event to be staged there. Both the Hook of Holland incident and the Queen’s Day Parade attack show that, just because an event has been staged somewhere before without incident is no guarantee that something unexpected won’t happen next time. Capacity It is quite right that capacity varies from base station to base station, dependent on the normal or foreseeable demands placed on the system in that area. However, as with coverage it should be planned to take account of demands likely to be placed on it. It is also possible to move capacity around the system, or temporarily upgrade it to deal with planned events. As with coverage however, when it comes to the spontaneous incident users must make do with what is available. Recognising that a surge in demand will place strains on any system, no matter how much capacity is available, what is to be done? Probably the easiest method of reducing demand is to restrict the number of talkgroups in use at the scene, and this will also assist with ensuring communications are all informed. This may impose operational, rather than technical, queuing; but at least the user is aware of what is happening, rather than being unable to understand why the system that normally works so well is failing when it is most needed. This is a combination of planning, knowledge of the system and training. In the Belgian model we see that users are trained to change talkgroups so that everyone involved in a spontaneous incident is, initially, on one talkgroup. At the scene they can be split into groups to undertake different types of work and spread out onto different talkgroups if required. An alternative may be for dispatchers to instruct users to change talkgroups at the time of, or prior to, dispatch. Geographic restriction of talkgroups through the use of valid site profiles is one possibility, which can prevent the ‘dragging’ of talkgroups into areas where they are not required to work.

2

Training of users to explain why it is important that they do not monitor their ‘home’ talkgroup when they are otherwise engaged is another. There is also a need during crisis situations for someone with system knowledge to establish quickly where the users are registering on the system, how much capacity is available there and whether the system is coping. If it is not coping well, decisions must be made as to what can be done to reduce demand, and those decisions must result in immediate action being taken to rectify the problems.. This may include denying the ability to make telephony calls, restricting the use of private calls or in extremis stunning radio terminals that are monitoring unauthorised talkgroups. Fleetmapping It is important that agencies recognise that fleetmapping is the process of making the radio system fit the needs of the organisation through the use of priorities, permissions and settings and that the person doing this requires extensive knowledge of the organisation, its size and shape and its methods of working, policies and procedures. It requires some knowledge and understanding of the system and what it will and won’t do, but it is an operational rather than an engineering task and must encompass not only the system settings but extends to control room equipment and radio terminals; the latter should be configured in such a way as to primarily make life easier for the field operative, rather than the maintainer or programmer – though their needs need not be mutually exclusive. This must also reflect the needs of interoperating with other agencies, and indeed those parts of one’s own agency with which one does not ordinarily interact. Control Room Staff employed in control or communications rooms are, by definition, communications professionals and should have far more knowledge and understanding of the TETRA system – ideally including being in a position to offer advice to operational users on the most efficient use of the system and on the radio terminals. The control room operative will have access to far more technology than the average field operative and could have access to all of the event and incident plans, coverage and capacity statistics and radio operating instructions. The training of key communications room staff as Tactical Communications Advisors (also known as ‘radio communications officers’) would ensure that they are immediately available 24/7 to offer advice to operational commanders charged with dealing with an incident. As complaints about the radio system are usually initially addressed to the control room it would make sense for there to be a system of liaison between the organisation and the network operating staff here, and this as part of business as usual would lead to better liaison at times of emergency action being required. It must also be borne in mind that the more talkgroups that are used may impact on the number of control room staff required. So the discussion must be held - how many people do we need in a control room during a normal working day and during a crisis. The solution for handling more communication seems to be more control room staff?

3

Operational As with the lessons learned section, there are several interconnecting items covering planning, training and exercises. Rather than repeating them, let us compare 2 of the Netherland incidents. In the Turkish Airlines crash we had a spontaneous incident where resources from police, fire and ambulance were drawn in an unplanned fashion from across several safety regions and dragged their home talkgroups to the scene of the incident. This caused huge problems with the system quickly becoming congested and people feeling they needed to resort to other communications media in order to communicate with the control room. In the case of the Queen’s Day Parade at least two of the agencies, police and ambulance, would have been present for the planned event, monitoring a small number of talkgroups. When the planned event became a major incident the system had very few problems coping, because there weren’t large numbers of talkgroups suddenly migrating into the area and swamping the system. This equates to the Belgian model where users have already moved away from their home talkgroups in order to manage an unusual incident before arriving at a central location and demonstrates what a difference this makes. The closer to business as usual we can make dealing with a spontaneous incident the better; the practise of regularly changing talkgroups means this is not a big deal for the Belgian police, whereas elsewhere the users may very rarely be required to change talkgroups and so are unable to remember how to do so – or they don’t know where in the radio the required talkgroup can be found. Specific to Netherlands, the following comments have been made during the workshop; although they were made by Netherlands staff they are equally valid for many TETRA projects. As has been said by various senior officers: “the C2000 system is the best we can get”. Don’t close down your Tetra project too quickly. Keep people available inside the public safety organisations to deal with mobile communications. Training and exercises must be done on an ongoing basis. The control room must, in communications terms, be in the lead during the incident. Good communications are required between the control room and the maintenance organisation. The Fleetmap is the outcome of the organisational and communications structure. Standardisation of working processes are required, so end users know what to do and how to react in a crisis.

4

Technical There are those who advocate the use of mobile base stations to overcome limited coverage or capacity; in a situation like that experienced in Cumbria in the UK where the major incident phase lasted for several days, this may be fine. But if we look at the Turkish Airlines crash, normality – in terms of radio resources – had returned more or less after 4 hours or so. It is unlikely that a mobile system could be transported and configured in much less time than that. Not that there is necessarily no place for mobile base stations; as has been said, in a long term incident where a disaster has affected the normal infrastructure – or for a planned incident where there is no coverage normally, fine. In London the Metropolitan Police have used an ‘overlay’ system to give extra capacity during planned events, but this takes a great deal of planning and preparation and is of no use during a spontaneous incident. Some people have asked whether there couldn’t be extra capacity in each base stations, but disabled until it is needed. Unfortunately this still requires more equipment, which has to be paid for whether it is used or not. In the early days of TETRA, DGNA was spoken of as the magic bullet that would resolve many communications problems. In fact its use is not as straightforward as was originally thought and it is not widely used. There are problems with different radio terminals reacting to DGNA in a variety of ways and with how users need to respond. Manufacturers are constantly striving to improve their products and provide something that defines them as being better than their competitors and it is inevitable that TETRA will continue to evolve as a result.

Chapter 7

CONCLUSION In the current financial climate, where everyone seems to be asked to do more with less, there is more than ever a need to ensure that best value is being extracted from available technology. Though operational commanders will argue that they cannot afford to release staff for training and exercises, the difficulties of doing so must be weighed against the certainty of criticism in the event of a major incident where communications are found to be poorly managed or understood. In the Turkish Airlines crash it may be viewed as fortunate that the loss of life was not more severe and that the aircraft came down in an area of open land whose clay surface absorbed much of the impact of landing – rather than crashing onto a residential area and exploding or bursting into flames. It can only be imagined that the communications system would, in such circumstances, have been placed under even greater stress. In the Hook of Holland incident police officers felt they had no option but to protect themselves by opening fire because they were attacked. Their stress levels were also increased by the fact that they were unsure whether the communication system (e.g. the emergency alarm function) was working. The upshot of the two incidents is that it appears police officers have had their confidence in their radio system undermined and the Dutch Public Safety organisations are now having to review the system, the fleetmap and officer training and may be compelled to install extra coverage and/or capacity. It is almost certain that the C2000 system will be found to be fit for purpose, provided that its users are properly trained and have an understanding of how the system is intended to work; confidence in the system could be restored by exercises held in each region showing that the system can cope with major incidents. After all, C2000 had been in use for a number of years prior to 2009 without its efficacy being called into question.