Generic Vehicle Architecture and the UOR Protected Mobility Fleet

Back in September 2011 I wrote a piece about Generic Vehicle Architecture, with the wider Land Open Systems Architecture, and how I thought it was one of the smartest things the MoD had done for a long time.

Click here to read the original post

Land Open Systems Architecture (LOSA)
Land Open Systems Architecture (LOSA)

Just to remind you, GVA, or Def Stan 23-09 is an open standard designed to place information at the heart of a vehicular system. The objective of GVA is to create a single, standard digital electronic and electrical architecture for UK vehicles

DefStan 23-09 defines physical and communications interfaces on a vehicle to allow interchange of equipment and provides definitions of the Human Machine Interface.

The purpose is;

The purpose of this Def Stan 23-09 is to enable the MOD to realise the benefits of an open architecture approach to Land platform design and integration, especially in regard to platform infrastructure and the associated Human Machine Interface (HMI) in order to improve operational effectiveness across all Defence Lines of Development (DLOD), reduce integration risks and reduce the cost of ownership across the fleet. This is achieved by mandating and applying the appropriate interface standards

With news announced at the recent Defence Vehicle Dynamics show that most of the UOR vehicle fleet (with the exception it would seem of Warthog) will be being bought into core, Selex had on display a modified Cougar with their GVA conversion equipment.

The solution replaces the existing crew displays with more compact, high-performance GVA HMI compliant displays and adapts the existing sensor suite to exploit the benefits of the GVA digital video standards. The innovative and modular software based approach implemented by Selex ES has also enabled a greater level of functionality to be embedded in the system compared with that which is currently fielded.

I think this is pretty significant because it allows the hugely disparate protected mobility fleet to at least have some degree of commonality on the electronics, thus reducing training overheads and maintenance costs.

As the ‘UOR into Core Implementation Strategy’ begins to kick in, vehicles return from Afghanistan and the longer term fleet composition becomes clearer this kind of commonality will become vital to controlling costs. It might cost in the short term to implement the GVA compliant equipment but the benefits will be enjoyed over the longer term.



Read Def Stan 23-09 at the link below


Newest Most Voted
Inline Feedbacks
View all comments
June 26, 2013 10:47 pm

Hmmph. Much as I’m happy there is *a* standard, there’s a lot here which seems either over or under specified. For example:-

7.2 and on

OK, vanilla ethernet and 10G over single mode fiber. Nice point about physical separation for differing security grades and being unable to span different ports. But:-

– nothing about the minimum number of fiber/copper ports to be made available
– nothing about the layer 3 device embedded in the bus. What are it’s limitations, particularly if it’s switching 10G? Will it enforce traffic separation via ACL’s?
– great to have a publish/subscribe and timestamp services, but how are the devices going to get IP addresses? Someone needs to define DHCP service
– Bowman gateways and the like are good, but you need to discuss what the gateway can do. Will it firewall, can it provide DNS?
– port security. I worry about these lovely buses, which anyone can plug into…in order to power their secret squirrel GPS/secure radio allowing someone to tag machines before they drive out of a PB. 802.1x is overkill and might cause some worries with regard to the RWS having the wrong certificate and refusing to boot!
But there are commercial switches that will disable a port when DHCP snooping shows no valid IP was assigned.

8 Power

Defines the types of ports that should be available. But:-

– no minimum power budget specified. Sort of essential if you are designing equipment :-(
– nothing about equipment cooling available or not, or the heat budget limitations for equipment.

Nothing about EMP either, although Tempest is listed

June 27, 2013 7:18 am

wf – I get the impression you understand all this computery-data-busery stuff. Someone has to…

When last I played with such an architecture it was the network’s responsibility, not BOWMAN’s, to firewall that interface. That may have changed in the years since.

Power budgets and environment control measures are down to the platform designers and, because building in the maximum foreseeable of both would pretty well take over all available volume, the platform designers will reach a compromise based on defined role in the requirement plus a modest growth element. So, wf the equipment designer, as ever that’s the subject of discussions with the platform engineers. Although of course your equipment offers supercomputer performance while drawing microamp levels of juice. Obviously.

Ref security – while I see the advantages in publishing generic interface standards in allowing suppliers to offer true plug&play equipment (optimistic, but hey…), it is akin to publishing the bank’s vault combination to enable anyone access to put money in without hassle. One person’s convenience is another’s opportunity to grab what’s not theirs. In IT systems I recall very heated arguments between security experts and network experts over the need for Master Passwords and power-up backdoors to crack into a system when Users are all locked out. Same sort of issues.

EMP has been quietly ignored over the past decade. Mostly I suspect because defence ministries the world over assume all future conflict will be exactly the same as the most recent one, and in a dozen years in the sandpit our opponents never came up with a credible nuclear device. Sadly, and despite much campaigning within projects by people like me, MOD has allowed EU rules and regulations to trump sound military engineering so for example all diesel powered vehicles must comply to emissions regs which means tight electronic engine control. The controllers are COTS. So on the first special going POP! the EMP will rip through the controllers of all the modern equipment with completely unpredictable results, although you’d guess most will not operate. And we’ve been busily fitting COTS GPS, COTS IT, even some COTS comms kit. So whether the sparkly GVA databus still works is almost irrelevant, isn’t it? But not to worry, nearly all other nations have similar electronics-laden COTS based kit so no-one gets an advantage…

June 27, 2013 8:11 am

: “When last I played with such an architecture it was the network’s responsibility, not BOWMAN’s, to firewall that interface”. The trouble with this, is it makes no sense that way. Each bus should in theory be connected to only trusted devices, so there’s no need to firewall anything. Only the Bowman device allows offboard access, so it should be doing the firewalling.

They should probably define which equipment is allowed to use Wifi/Bluetooth as well :-)

I am reminded that the likes of a SIM module would actually be quite reliable for granting access to the bus.

“Power budgets and environment control measures are down to the platform designers and, because building in the maximum foreseeable of both would pretty well take over all available volume, the platform designers will reach a compromise based on defined role in the requirement plus a modest growth element”

Doesn’t this completely void the concept of a generic architecture? We want people to be able to design to a generic standard, which can be fitted to any vehicle using GVA. Obviously some vehicles will have different power budgets, but as they have defined the power connectors, the logical extension would be to standardise classes of equipment, so they have standard physical slots in which to plug, each of which would have the standard plug (which is defined) and a limit to emitted hot air.

EMP: well, someone really needs to start taking this seriously soon. Countries may not want to use it (doubtful about that too) but it’s not just countries we have to worry about :-(

June 27, 2013 9:13 am

wf – interesting you bring up the term SIM. For many years I have discussed an idea with the likes of DSTL and QinetiQ for the creation of a ‘military SIM’ to be used with ordinary every-day mobile phones – a device that adds a key into the chirp so that a) civilian base stations ignore the traffic as corrupt, and b) military base stations only process the appropriately keyed data. I imagine the technology is already in use to separate different network users operating in the same band. With such a device the User Terminal is a cheap and cheerful throw-away handset. In areas away from home territory, mobile military base station micro-nodes would need to be set up – one suggestion someone I discussed this with came up with was to parachute in battery powered one-shot self destructing pico-nodes in the hopes the chute snags in treetops… I think I heard the encryption on GSM is already quite good – I have no idea if the military would be content to trust it (to low levels of security classification). But the functionality built into even the cheapest mobile would be leaps ahead of personal voice radio – over (cht!)

As for future-proofing generic interfaces so that new equipment can be built to plug seamlessly into a pre-prepared slot? Military platforms, especially those on the ground, have little spare room for neat empty equipment racks with piped in cold dry air, there just in case someone wants to add a standard form factor box later on. In any case I would be pretty confident that the GVA compliant equipment might be standard round the back, but the front will be a vipers nest of cables, connectors and controls to deal with all the stuff GVA didn’t accommodate. Such is life.

I am, I will admit, fairly unexcited about GVA. When first the idea for a standard network was put forward for ground vehicles it was CANbus. Then Ethernet 802. Last time I saw real data, Ultra were suggesting a fixed TDM protocol on fast Ethernet to remove the mass of “Does anyone want to talk to me?” polling messages. Life moves on apace; the ‘standards’ are likely to go obsolete much much sooner than any platform reaches OSD. So we would be faced with upgrade costs purely to keep GVA compliance even if the need for new equipment wasn’t a driver. Not efficient to my eyes.

And as a reminder I have many bits of ‘plug&play’ IT equipment stored away that is unusable because the standards have moved on – printer with parallel port, PCI cards, SCSI drives – and storage media no longer usable – 3.5″ diskette, even a few 5″ floppy disks and a cassette or two in the loft. And then there are software packages no longer usable because operating systems can’t be fussed to put in backward compatibility. All in all the commercial world’s ‘plug & play’ window is probably little more than 4 years long before the standards have moved on. In the defence world you’d only be halfway through the procurement process before the system bid would be out of date.

So – apologies for long post – the only way to hold GVA to a useful thing is if it is set and locked at a given standard and not updated with the commercial world. Then you could be sure that you could add a new standard box onto the green vehicle that’s just spent the past 20 years in a heated shed doing nothing.

But equipment suppliers would be stuck with an ageing interface no longer supported by the commercial IT market.

VGA is a good idea, but for the fact that military equipment is expected to serve for decades, the defence procurement cycle takes years if not decades, and the commercial standards move on a generation roughly every 18 months. I’m afraid I can’t see how the two streams (military and commercial) can ever be joined by a generic interface purely because the streams travel at such hugely different speeds.

June 27, 2013 9:56 am

I seem to have read somewhere that GVA is based on CAN Bus. This being the case it’s interesting to note that at Def Con in Las Vegas at the beginning of August there’s to be presentation on the security vulnerabilities in CAN bus. Opens all sorts of interesting battlefield possibilities.

June 27, 2013 10:04 am

Oops! When I said “VGA is a good idea” I did of course mean “GVA is a good idea” – dexlyxsia rules…

At a GVA briefing I attended I am pretty sure CAN had been dropped as too slow and constrained – certainly for multiple streams of concurrent high def video or the like – and Ethernet of some colour was the new standard.

June 27, 2013 10:54 am

Chris, now VGA, that is retro. :)

For commercial and military nodes, is having so many of them a good idea? Even if the civilian nodes have a different protocol, there is still node interference problems, and the more you have, the worse it becomes.

June 27, 2013 11:10 am

: yes, you can see a lot of use for a hardware security module (which is effectively what a SIM is).

I disagree with your pessimism with regard to GVA, primarily because I have seen everything in the day job converge on ethernet. TDM voice, security cameras, door locks, storage networks…the lot. If you have single mode fiber, you are effectively future proofed very nicely, thanks. No requirement for rats nests anywhere :-)

With regard to “standard slots”, everything here has converged on the 19 inch rack. Most vehicles won’t need “compute nodes”, just the computing equivalent of an iPad which can be slotted anywhere. If the bleeps need more, fitting a compact 19 cabinet isn’t hard. Job done

June 27, 2013 12:46 pm

Dont we run into problems with obsolecence. If we have a standard for the mllitary it will havr to last 20 30 years to makes sence from design to replacement.

Try looking at the electrics on a mk3 ford escort . Or Alllegro for comparison. Or yout 10 year old mobile compared to your Android!

BTW yea to whoever made the emp point. Electrics and deisels should not be mixed.

June 28, 2013 1:31 am

, no you don’t really have any problems with obsolescence on the timescales we are talking about. If we were talking about this in 1983, we would probably be discussing the utility of deploying ISO HDLC over RS-422 as a data standard; and, would probably have come up with something along the lines of MIL-STD-188 (1988). This combination provides multi-point data comms at up to 10Mbps, more than adequate for 20yrs, and easily upgraded to higher speeds to give you your 30-40yr horizon.

Btw, as a PDA, my 15yr old palm pilot was in many respects a superior piece of kit to my modern android phone and certainly more rugged, and as a phone, my 25yr old handset was far superior to my android in terms of power consumption, range, and reliability, attributes that are of greater military significance than comparing Nokia ‘snake’ with an Android’s ‘angry birds’.

I see this quite a bit, even here. People assume that the pace of innovation they see in the bells and whistles of commercially driven ‘obsolescence by design’ CE kit should be matched by other more pedestrian and pragmatic technology areas. It is much much harder to invent a revolutionary solution to pragmatic problems then to design a revolutionary new ‘style’ to exploit the next evolutionary increment in the number of pixels I can fit in my hand. In a race between the laws of physics and the laws of aesthetics, aesthetics will always be moving ‘faster’.

If you wire a vehicle with fibre capable of carrying (today) 1-10Gbps, you will have plenty of headroom for the next 20yrs, and with incremental upgrades to the spec, for the next 40.

June 28, 2013 1:59 pm

@Andrae: yes, there is indeed scope now for data and power standardisation that will actually stand the test of time these days. With regard to a lot of the electronic kit, we’re going to have to accept a “when it breaks throw it away” attitude more though.

For an electrical engineers out here: why aren’t we using AC now? Why stick with DC?

Steve Rodgers
Steve Rodgers
June 28, 2013 2:37 pm

Fleet systems are very important in the world of transportation. Tracking has become a part of our day in every other area. Why not transportation too. They make transportation systems much more efficient for the public as well. Thanks for the post!

June 28, 2013 3:38 pm

wf – ref “why aren’t we using AC now? Why stick with DC?”

I think the key factor is generation. While an alternator (as fitted to all modern vehicles) is an AC device, it is not frequency controlled – the raw output has a frequency proportional to shaft speed. I suspect for efficiency purposes the regulation is mark-space switching of the field coils and not analogue which would chop and hack the output waveform horribly, but I might be wrong there. It doesn’t matter in this case because the power is rectified immediately – beyond the alternator casing there is nothing but DC. Generating a clean constant-frequency waveform typically requires a constant shaft speed – on passenger jets which have 400Hz three phase AC requirements, the generator was attached to the engine through a constant-speed drive (CSD) which was a heavy complicated and brutally expensive unit packed with variable displacement hydraulic pumps and swashplate motors. Yeuch! It is possible by now that the AC is created electronically – when I last played with such things the first few digitally synthesized AC power units were being advertized. Inverters have been around for a long while now but have never really gained much favour. So much so that for one project where AC was required, the answer was to drive a normal AC generator (big) off a DC electric motor that was speed controlled (bigger).

AC batteries are really hard to find too….

June 28, 2013 4:08 pm

“AC batteries are really hard to find too.”

… you damn joker. :P

Or at least I hope it is a joke, otherwise, I’m really behind the times.

And please don’t say an AC battery is a generator with a tank of gas…

June 28, 2013 4:49 pm

Obs – you’ll find the AC batteries in the company Stores on the same shelf as the left-handed screwdrivers, elbow grease and spare holes, and just round the corner from the curved drills. Really! Fancy not knowing about AC batteries!

June 28, 2013 5:28 pm

So that means they are on the same shelf as the AA and AB sized batteries and just before the C sized right?

June 30, 2013 6:30 am

Bovington snippet No.2:

Talking with one of the Tank Museum drivers, he commented he’d read somewhere that Tornado’s fighter capability would after EMP be about 20% of the capability of a Spitfire after the same. Spitfire’s magneto ignition would probably survive, the instruments are mostly mechanical and those that use electrons are basic electric meters, the radios using glowing thermionic valves (tubes for our American readers) would be quite robust. Tornado on the other hand has complex microelectronic controls for everything. I would hope they are robust against EMP if not immune to it, but all the same it does highlight how much of a vulnerability we have wilfully built into modern equipment in the desire to add more and more high tech wizardry as ‘force multipliers’.

Best we build a few squadrons of Spitfires though. Just in case

June 30, 2013 7:05 am

Chris, anything boxlike and metallic tends to act as a Faraday Cage, which makes EMP much less effective than it could be.

Planes unfortunately have other design considerations other than EMP resistance as their main goal, though they are slightly hardened to prevent electronic interference between equipment, not to endure a mid range nuclear strike.

June 30, 2013 8:40 am

Obs – agreed, but for two points; firstly that Faraday Cages need to be complete to work, so the likes of hatches (ground vehicles) and canopies (aircraft) tend to leave big holes for the vicious magneto-electrons to climb in, and secondly while it is pointless to protect a flying machine from EMP aspects of sizeable nuclear blasts (the pressure front would just swat them from the sky) there are devices I have heard that are small in blast effect but still whack out a serious EM pulse. And this is assuming that the secret squirrel weapon establishments around the world are not busy perfecting their “focused EMP” projectors…

I am not a Luddite (well not always) but I see advantage in at least some military kit being basic, robust, repairable by squaddie with string gaffa-tape & wire, and insensitive to soft-kill factors that affect computerized gubbins. In particular and related to GVA I have made direct request of the MOD SE lead that the multi-layer architecture retains at the lowest level a simple battery-switch-load DC electrical system that can be fault-found and repaired in the field without recourse to an IT helpdesk or software weenie.

And as a point of balance between military need and eco-protection regulations, the modern rules state EU regs must apply and diesel packs must be up to the latest eco-standards to protect the environment. This may render them unpredictable and dangerously self-willed (like engines shutting down because sulphur is detected in the exhaust – exactly what you need when extracating self from incoming fire). One FV432 sold by Witham recently had covered what was believed to be a genuine 39 miles under its own steam since entering service in the 60s. Just how much damage can be done burning 10 gallons of diesel badly? And how much eco-damage is done manufacturing all the control electronics & sensors, and mining all the fancy minerals used in cats and plating? When I worked at Alvis I was quite shocked that front line combat vehicles might average as much as 1000 miles a year (logs & general duty vehicles much more, obviously) – if a full long service life is measured to be an average of under 40,000 miles, you would have thought a pollution waiver might be sensible in order to ensure they remain fit for battle no matter what fuel they are fed, or what EMP might occur. Bring back mechanical metering, I say…