8/8/2019 EN-014-04

1/16

4-l

AGING AVIONICS- A SCIENCE & TECHNOLOGY CHALLENGE ORACQUISITION CHALLENGE?John Ostgaard, Juan Carbonell, Stephen BenningAFRL/IFSC2241 Avionics CircleWright-Patterson Air Force Base, Ohio, 45433-7334, USAThe fleet of aircraft which we have today,represents 90% of the aircraft which we will havewell into the next century. With defense budgetsshrinking, fewer new aircraft will be purchased.Avionics modernization is the highest leveragedapproach to maintain and increase capability. Thistreatise discusses the problems associated with our

existing fleet of aircraft, both from a maintenanceand modernization perspective. Examples ofobsolescence and retrofit issues are provided tohighlight the seriousness of the problem.The science and technology challengesare the focus of a small set of technologies aimedat reducing the cost of our aging systems. Thesecosts include maintenance as well as upgradecosts. Aging systems are different from newsystems because there is an existing infrastructure.It is much easier to start from scratch than

modify something already in existence. Specifictechnology areas that will be discussed include:a. Avionics softwareb. Group A Modificationc. Avionics Maintenanced. Application of COTS productse. Obsolescence/Environment for RapidDesign ChangesAcquisition related issues have to beaddressed in consonance with the Science andTechnology challenges. Issues that will bediscussed include: Open systems, plannedtechnology insertion/planned obsolescence,verification/validation, warranties/repair strategiesand organic/contractor support. As part of thisacquisition discussion, industry pursuit of aproduct line investment strategy will be addressedas their adoption of best commercial practicescontinues.

A possible Avionics S&T strategy will bediscussed which outlines an approach for infusionand transition of advanced technologies into theexisting fleet given an acquisition friendlyenvironment. Discussion on modeling, simulation,and prototyping will be provided to furtherexplain the fleet wide application of thesetechnologies.Background

Today DOD is being challenged to domore with less. This has focused attention onaffordability opportunities and challenges withemphasis on commercial or commercial-basedavionics and electronics. The life of older(legacy) aircraft is being extended while theiravionics systems are becoming obsolete, and moredifficult and expensive to maintain and support.New aircraft also need to be more affordable andhave functionality and performance equal to orexceeding that of existing systems. Lowerupgrade and support costs for ex isting aircraft andlower acquisition and support costs for nextgeneration avionics systems, are primarychallenges.

Our Nations aircraft are capable, buttheir electronics are old and aging rapidly. In thepost 2000 era, these aircraft will experienceworldwide deployment in support of our activefront-line forces. The majority of our Nationsfleet of aircraft through 2020 will be legacyaircraft as shown in Figure 1.

The tactical fighter fleet is aging rapidlyin large numbers, with the projected average ageof all fighters in the USAF and Navy inventoriesestimated to be over 20 years by the year 2020.

Paper presented at the RTO AVT Lecture Series on Aging Engines, Avionics, Subsystems and Helicopters,held in Atlantic Ciq, USA, 23-24 October 2000; Madrid, Spain, 26-27 October 2000, and published in UT0 EN-14.

8/8/2019 EN-014-04

2/16

4-2

25

0199e 1994 1996 19% 2XJ XW 2034 2% 208 2010 a312 2014 2316 2018 2@0

Figure 1 Tactical Fighter InventoryClearly, avionics on these aircraft willrequire modernization during the next decade ifthese weapon systems are to maintain theirreadiness status. Future aircraft modernizationwill focus increased capabilities which requireeffective communication with active forces,concurrent use of weapons, compatibility withelectronic combat systems, up-to-the-minuteaccess to off-board intelligence, an effectivemeans of cooperative targeting, high accuracybombing with inertial guided weapons, and on-board training for pilots and ground personnel.These capabilities must be inserted affordably andwithin the cost or schedule constraints imposed bythe military budget.Many technological advances havingprofound implications for military capabilities arethe result of research and development conducted

by commercial firms for essentially thecommercial markets. Digital products found inpersonal computers have become the center offocus on commercial technology insertion as theyhave become much more powerful than many ofthe computers being used in our weapon systems,and at dramatically lower prices (see Figure 2).The power of the processors gives intelligentweapon systems their capabilities. The combinedfactors of expanded capabilities and lower costsare clearly compelling reasons to integrate these

commercial technologies into our existing andadvanced weapon systems.

Legacy aircraft issues and reasons for upgradeRecent DOD budgets and the politicalclimate are forcing the United States militaryaircraft to remain operational well beyond theirprojected service life. These life cycle extensionshave prompted many legacy aircraft upgradeprograms. In addition, todays military aircraftmust perform an increasingly broad range ofmissions from tactical/strike to reconnaissanceand electronic warfare. These emerging missionsencompass a wide range of avionics functions,performance, and cost. Several platforms are inneed of upgrades due to the need for increasedconnectivity, the lack of spares, and directedinitiatives that require increased avionicsperformance/functionality. However, currentbudgets have been reduced to the point where it isdifficult to meet all requirements using thetraditional upgrade methods employed during the1980s and early 1990s.In the past, piecemeal systemenhancements were performed to meet newrequirements. These piecemeal retrofits haveresulted in point designs that only support aparticular problem and platform. As a result,

8/8/2019 EN-014-04

3/16

4-3

other platforms could not readily leverage thesedevelopments. Many of these upgrades were notapplicable to subsequent upgrades of the samesubsystem, and as time passed, upgrades werelayered upon each other compounding themaintenance problem. Tight coupling of thelegacy systems forced the major parts of theavionics system to be recertified, which became amajor cost driver for the upgrade. Furthermore,these upgrades must be performed and scheduledinto the existing maintenance schedule or theexisting block upgrade cycle.

upgrades are currently funded. Some issues arestructural in the present procurement andmanagement approaches. Decisions are made ona program by program basis, and we see no trendto decrease the authority of the system programdirectors. Any changes to aging aircraft avionicswill have to make sense to the weapons systemprogram director, while providing capabilities andsupportability acceptable to the user command.Introduction of a new supportability concept, suchas one- or two-level maintenance, will have toprove itself in the context of the existingsupportability plan for the weapon system,

Figure 2 Commercial and Military Processing Technology ComparisonCommercial digital technologies haveincreased in performance several orders ofmagnitude over the past decade. Advanced da taand signal processing capabilities presentopportunities for extending the life expectancyand meeting new mission requirements of ourexisting military aircraft by inserting newcapabilities in weapon system platforms.Beyond commercial technology insertionand basic logistics problems, there are issues ofadministration, funding, and organization of how

possibly with a 3- to 5-year pay-back.Reasons to upgradeNew and enhanced mission requirements.Legacy avionics designs are generally based ontechnology that is at least 15 years old and at least2-3 orders of magnitude slower than commercialequipment available today. Thus, legacy designsdo not have sufficient capacity to meetrequirements demanded by the new and enhancedmissions, even if significant spare and growthroom is built-in.

8/8/2019 EN-014-04

4/16

4-4

Increased reliability and better fault isolation.Upgrades using modern technology allow forincreased reliability because fewer parts and fewerinterfaces are required to provide the samefunctionality, thus providing L ife Cycle Cost(LCC) savings potential. This can also lead tobetter fault isolation because fewer can notduplicate (CND) anomalies are produced.Component obsolescence. The military qua lifiedcomponents used on legacy systems are becomingobsolete at an increasing rate. This provides anopportunity to upgrade the system using COTSparts and the potential for substantial LCCsavings.Avionics Software

Modern avionics systems are expensive toupgrade because the software is designed andbuilt to tightly integrate and operate on customhardware. This situation was driven by the needto squeeze every last cycle out of hardware toimplement that next feature. Simply stated,current avionics software is not portable.However, software upgrade and maintenancecosts cannot be studied or solved without thedefining context of a specific system architecture.Given this situation, major software cost driversfor the upgrade of legacy systems include thefollowing:1. Recovery and understanding the originalsystem requirements and designconstraints. Lack of understanding ofthese requirements causes rebuilds andsometimes program restarts due toerroneous assumptions and poor designchoices based upon incomplete orinaccurate information.2. Interfacing with the legacy designphilosophy of the avionics system andsub-systems. This applies for bothhardware and software (language) issues.3. Interfacing to legacy system operatingsystem environment to meet bothscheduling and communication interfaces.Often times this is not a clean interfacedue to the cyclic nature of most deployedsystems and the hardwiring of addressingon the communications bus (i.e. 1553).Traditional approaches to solve theseproblems (e.g., sharing memory, bussnooping, etc.) are error prone because the

new system cannot determine the absolutestate of the legacy system.4. Re-certification and Re-validation of thesystem and/or sub-system as they areincrementally inserted. The cost ofvalidating the software and hardware forsafe and correct operation each timefunctions change will dominate the cost ofany upgrade involving functionalmodification.5. Fitting major software/hardware buildsand insertion into the normal maintenancecycle. The time to develop and insert anew avionics suite (or a major sub-system) exceeds the standard maintenancecycle.Upgrading and re-hosting software from alegacy avionics computer to an upgraded one istraditionally a very expensive effort. Thesoftware must behave substantially the same onthe new processor as on the old one, so interfacerequirements must be captured and tested. Muchlegacy code is closely dependent on thearchitecture and operating system of the legacycomputer and is often written in an obsoletecomputer language. These problems conspire tomake direct re-hosting of the code nearlyimpossible. Several approaches, includingreengineering of the code or encapsulatingportions of legacy code with modern wrappers,show promise by retaining some of the investmentmade in the original code development. Otherapproaches, such as translation and emulationretain even more of that investment at the expenseof future maintainability.Standards apply at all levels ofsystem/software. To aid in the classification ofstandards, the Society of Automotive Engineershas a reference model called the Generic OpenArchitecture model (GOA). This referencemodel, shows a detailed breakout of key interfacesfor systems. Each of the GOA interface areas areopportunities for reducing the level of interfaceincompatibility within future systems, improvingupgrade opportunities, and providing the basis forprocurement guides. For example, IEEE POSIXdefines interface standards at the GOA level foroperating system and operating system services,and has many commercially availableimplementations. POSIX allows any applicationsoftware to be portable to another system that usesa POSIX solution for the level of interface.

8/8/2019 EN-014-04

5/16

4-s

Widely commercially used standards (e.g.CORBA or POSIX) provide access to a widerbase of professionals that have used the standard,and the implementations of that standard havemore operational hours-i.e. fewer defects. Thiscommercial appeal also provides better andgreater numbers of available developmentenvironments. All these factors lead to lower costfor systems that can take advantage of commercialsoftware interface standards.Software development cost has been thesingle major cost driver in most avionicsimprovement programs. An effectiveincremental upgrade must utilize COTStechnologies to shorten the software developmentcycle and increase productivity. For MIL-STD-1750 Processor-based software development, fewdevelopment tools are available, especially forlegacy programming languages. Furthermore, itis not uncommon for legacy codes, based on olderlanguages such as FORTRAN and JOVIAL, tobecome difficult or impractical to maintain asfamiliarity with the target hardware and highlyoptimized control and application programs andtheir development tools diminishes when theengineers who designed them move on. For thesereasons, software engineers have begun to migrateaway from proprietary, application-specific

architectures towards well-defined COTS opensystems standards and technologies wheneverpossible. (See Figure 3) This involves using

system environments like DOS, UNIX, VTRX@(MicroTech), VxWorks@ (WindRiver); andinterfacing to software Application ProgramInterfaces (API) like Windows@ (Microsoft) orTCP/IP communication protocols,

Group A modificationGroup A modifications involve permanentmodifications to flight equipment on the aircraft.Avionics upgrades usually require major Group Amodifications at significant cost. Studiesconducted by the Modular Avionics SystemArchitecture (MASA) Program on the F-16 fleetindicate that at least three black box systemsmust be integrated into one before modular

upgrades can be affordably integrated into a singleblack box replacement. During this study effort,conducted in 1990 for the F-16, it was concludedthat the cost of additional Group A wiring changescould be justified with ou t-year life cycle costsavings. But as effective service life decreases,the life cycle cost savings are amortized acrossfewer years. In general, the cross-over point forhigh rates of aircraft modernization isapproximately 10 years of remaining service life.For low rate or low number modernization efforts,Group A development cost is assumed to be thedominate cost inhibitor. In various studiesconducted, over twelve change proposalsincorporating Group B equipment were altered or

Figure 3 Open Software Architecturestructured, well defined languages like C and reduced in scope to accommodate limitations ofC++; employing object-oriented programming the Group A changes. One significant programmethods; working within familiar; broadly was canceled due to estimations for F-l 6 Group Asupported, hardware independent operating cost impacts. The results of this analysis

8/8/2019 EN-014-04

6/16

4-6

provided an important assumption for aircraftmodernization. If future upgrades are small andincrementally funded, the typical upgrade may beprojected as a black box one-for-one replacement,referred to herein as in-place upgrade. As GroupB black boxes or LRUs are upgraded, the physicalwiring, cooling, power, and mechanical interfaceswill largely be left unchanged for the remainingservice life.

As part of avionics upgrades,modifications to existing aircraft electrical andmechanical configuration items may be requiredin order to interface with the new elements.These include power, cooling, wiring harnesses,mounting fixtures, cables, connectors, etc. Themodifications can be extensive and timeconsuming when redesigning of existing LineReplaceable Units (LRU), replacing and addingnew wiring harnesses, and on aircraft systemintegration and testing are involved.

A systematic approach that allowsinsertion of improvements utilizing wiring,installation, and other existing aircraft resourceswill drastically reduce Group A modificationcosts. The practice of using field installablemodification kits will further reduce time-to-fieldand unscheduled aircraft down-time.In addition to electrical and mechanicalchanges, a major consideration to upgradinglegacy avionics is their network architectures.MIL-STD-1553B Multiplexed Bus structure hasbeen the standard for avionics multiplexedcommunication implementations. While thedemand for faster computational capability anddata communication of high volume, intermixeddata types continues to increase, bridging theMIL-STD-1553B time division multiplex bus-based avionics architecture ( 1 Mbits/set.) into

one that utilizes several orders of magnitudehigher network throughput, such as SerialExpress, Asynchronous Transfer Mode (ATM), orGigabit Ethernet, has become necessary toavionics modernization. By expanding the use ofexisting MIL-STD-1553B cabling in conjunctionwith a scalable protocol implementation, Group Amodification cost can be minimizedChanges to the cooling infrastructure areexpensive and usually require changes to theaircraft structure. Thus cooling air found in most

legacy aircraft is the heat removal medium. Edge

conduction modules have a higher thermaldissipation capability than convection cooledmodules, plus they have better vibration andshock performance due to their central heat sink.It may be possible to use liquid edge conductioncooling with heat exchangers to the cooling air incases where size restrictions and the resultantthermal density of the modules require greaterthermal dissipation capability than cooling airprovides. In such cases, the liquid coolinginfrastructure (circulation and heat exchangers)may have to be self-contained in the avionicsupgrade. However, a prototype avionics liquidcooling system for currently fielded aircraft andfor the retrofit of state-of-the-art avionics has beendemonstrated. The electrically driven vapor-cycleheat pump was integrated into the F-16s currentcooling system and increased the cooling capacityby 5KW thus enabling the aircraft to beincrementally updated well into the nextmillennium with advanced, reliable avionics atlow cost.Avionics Maintenance

Typical Operations and Support (O&S)costs run approximately 70% of the total LifeCycle Cost (LCC). Cost distribution is spreadover Support Equipment, Depot MaintenanceEquipment, Personnel, Initial Training, InitialTechnical Data and Initial Spares. To reduce O&Scosts, avionics systems must be highly reliableand easy to operate. Timely repair orreplacements will minimize Mean Time To Repair(MTTR) and increase Mean Flight Hours BetweenUnscheduled Maintenance (MFHBUM). Legacyaircraft avionics upgrades must introduceinnovations such that system and subsystems leveldiagnostics can be performed to provide accurate,thorough problem identification and isolation. Inaddition, periodic preventative maintenance canbe automated and prognostic data made availablefor trend analysis. Hence, non-repeatableanomalies can be eliminated.

Can Not Duplicate (CND) faults andRetest OK (RTOK) problems account for alarge percentage of the maintenance events onlegacy aircraft. Better diagnostics will help reducethese problems, but some of the solution mustcome from the design of the system. Reduction inthe number and complexity of interfaces willincrease reliability and reduce CNDs. ThoughCNDs should occur less frequently in a reliableCOTS-based system, when they occur, a method

8/8/2019 EN-014-04

7/16

4-7

must be provided to save the system state in orderto duplicate the situation and attempt to diagnosethe cause of the CND in a laboratory.Mainstream commercial processors andsystems do not normally have much diagnosticcapability other than at system start time. Somesystems are capable of isolating memory errors toa specific failed memory module and someoperating systems can probe devices to determineif they are operating normally, but other than that,very little isolation is available. More expensivecomputers designed for high availability, such asservers and supercomputers, have some built infault detection and isolation along with redundanthardware to help tolerate the faults. Finally, errorcorrecting/detecting memory controllers can be

used to help correct for and diagnose failedmemory components. These techniques and othersneed to be investigated to determine howcommercial technology-based avionics upgradescan diagnose and recover from failures and aidmaintainers in determining where the failureoccurredDistributed avionics upgrades requiresystem management software to work with thehealth monitoring software on each processor tomaintain a consistent and correct view of the

health status of the system. When failures occur,the system manager must activate the programsimpacted by the failures on other processingresources. In addition, enough system state mustbe available to these newly-activated programs sothat they can begin operation rapidly and notcause any mission-critical functions to fail forlonger than a specified time. An overall systemhealth maintenance and reconfiguration approachmust be integral to the upgrade architecture and itsassociated system software.Critical to legacy aircraft processevolution is the support strategy and maintenanceimprovements introduced to legacy aircraftweapon systems. The definition of this supportstrategy is key to establishing lower operationalcosts when service life decreases as a function oftime remaining in service life. In addition,interfacing the weapon system to the globalnetwork is a major goal for future product supportsystems in order to establish a complete networkresource to determine combat status of individualweapon systems for local theater commanders

during weapon system deployment. The

technology provides users with automated pilotdebrief, advanced diagnostics and electronictechnical manual data presentation capabilities.The new technology is being used for currentlegacy aircraft as well as future programs.The Integrated Maintenance InformationSystems (IMIS) technology is a multi-levelsecure, integrated, distributed task and decisionsupport system that automates numerousprocesses to improve pilot debrief andmaintenance performance functions. IMISreduces management, supervisory and technicaloverhead by providing:

1)

2)3)

Real-time access and interface with customerLogistics Data Systems as well as otherDepartment of Defense external systems,Aircraft advanced diagnostics capability, andDisplay of Interactive Electronic TechnicalManuals (IETMs).

The system reduces the hours needed to service,troubleshoot, and repair all aircraft systems andassists in the achievement of the maintenancemission goals of high sortie rates, to minimizeaircraft down-time, and to provide maintenancewith minimum support resources. Obtainingaccurate data of aircraft availability and status iscritical to mission planning and repair operations.Achieving timely and accurate data from fieldoperations is an important aspect to understandingwhen and if weapon systems requiremodifications to be effective in their missions. Ascommercial technology is inserted into legacyaircraft systems, it becomes crucial to monitor thefield operations of these equipments, as criticalbuilt-in-test and self-test hardware and software isgenerally not built into commercial gradecomponents and systems. Hence the logisticsystem must be modified to accommodate andcontrol the infusion of technology into the fleet ofaircraft and provide a means for technologyrefresh, as shown in Figure 4.Aircraft data is received through a data transfercartridge (DTC) or via an aircraft maintainervehicle interface (MVI). Although capable ofoperating as a stand-alone system, the IMISinterface can communicate logistics data througha network connection to external systems. IMISconsists of a networked computer system that

8/8/2019 EN-014-04

8/16

4-8

Commercial Industry ilaLBr COTS Products ttz:!W ? l Designs^ ,_ $y

Defense Industry $l Industry Surveys & Analysis !Ij$l Industry Studies I Verification

Military Services $SUDDOrl

-Environment Il Products l Ruggedization (if Necessary) * Integrated Diag&tics

l Legacy HkSWl Readiness Needs

l Operational Needsl Technology Improvementsl Fleet Commonality l M . -.,-,-.-...-l Open System Standardization 1 -Functionality q

- Gnd Support Equipment- Operational SWlHW I.

-Service Warrantiesl COTS Products ? /

MaintenanceDemonsrrar/on HWSW l WarrantyExpires

Figure 4 Logistic Support is a Vital Aspect o f Technology Refresh

communicates with Maintenance InformationWorkstations (MIW) or Portable MaintenanceAids (PMA). The system components can betransported, allowing use during deployment.Current IMIS development efforts add IMIScapability to provide the system user with real-time access and interface with customer LogisticsData Systems as well as other DOD externalsystems.Anticipated benefits of the IMIS systeminclude:50% reduction in pilot debrief time30% improvement in troubleshootingaccuracy30% reduction in Re-Test OK (RTOK)rates40% reduction in the time to repair amalfunction or perform maintenanceactions resulting in:Improved weapon system availability andreadinessReduced aircraft down-timeImproved aircraft sortie rates

l Reduced weapon system life cycle costl Improved maintenance data collection

Currently, IMIS is in F-16 field serviceevaluation and supporting the F-22 Raptoraircraft. Further evaluation is required topotentially add a commercial interface standardsuch as Ethernet to allow direct access to aircraftweapon systems, either w ired, spread spectrumRF, or high speed IR links are being defined as anopen systems approach to field service andsupport. Another element of the IMIS system isthe integration of Interactive Electronic TechnicalManuals. The Interactive Electronic TechnicalManual system allows the user to locate requiredinformation faster and more easily than thecurrent paper technical manuals. IETMs are easierto understand, more specifically matched to thesystem configuration under diagnosis, and areavailable in a form that requires much lessphysical storage space than paper. Interactivetroubleshooting procedures have been added usingthe intelligent features of the IETM displaydevice.Application of COTS products to upgrades

The use of commercial technology-basedcomponents is becoming more and more essential

8/8/2019 EN-014-04

9/16

4-Y

for avionics upgrades because of the lack ofavailability of military-qualified parts. In addition,commercial components are normallysignificantly less expensive and have highercapability than similar m ilitary parts. Whilecommercial components are not designed fordirect application to military avionicsapplications, it is expected that methods can beemployed that will allow them to operate reliablyand cost effectively within avionics environments.The use of COTS components presents otherissues that need to be resolved in order to providea smooth transition to COTS-based avionics, andrecoup the cost savings that would make thetransition worthwhile.Due to the differences in theenvironments of the various legacy platforms, andthe resultant variations in requirements, aprocurement strategy must be developed that canmaximize the benefits of applying COTS to thevarious platforms. In addition, aircraft will beupgraded in blocks, but the COTS technologyused will most likely be obsolete before the blockupgrade is complete, so there needs to be aconfiguration management scheme andcontinuous upgrade process in place that allowsinexpensive insertion of newer components withlow-cost revalidation. Methods for certifying the

aircraft fleet with the various versions of the blockupgrades must be defined. Integration of theCOTS-based avionics with the legacy avionicsneed to be addressed; examples of integrationissues include, but are not limited to, interfaces,EMI, and architecture changes. Other areas, suchas support equipment and training, will alsorequire attention in order to minimize the costimpacts of the ever-changing COTS technology;this would include such areas as documentationchanges as well.Due to the fact that all aircraft in a fleetare not upgraded at once, some of them will havedifferent hardware configurations than others.Various methods exist for configuring theavionics hardware, both at installation and atmaintenance time. One possible techniqueinvolves having the avionics system perform aplug-and-play type operation of determining thehardware components it has and loading theappropriate software from a mass storage device.Another aspect of configuration control is

the logistics aspects of the hardware components.

It is possible to configure the aircraft in terms offunctionality as opposed to part numbers.Hardware components are made to be Form, Fit,Function and Interface (F31) identical. Thecomponents are then tracked according to their F31classification and not their part number. Aircraftcan then be made to be identical as far asconfiguration control, and the logistics can beimproved due to a reduction in part numbers(F31 numbers). Of course, each component willhave to meet some minimum requirements andtesting to achieve its F31classification.Mil-Std-490 has been canceled andreplaced with a handbook. Government / IndustryData Exchange Process (GIDEP) interaction withjust commercial components needs to be fleshed

out along with RADC traceability and controls.Mil Std 965, Parts control program, beingcanceled has be to dealt with philosophically. Isthis just a subcontractor issue? Are we partners?The current trend of higher clock speedsand shorter rise and fall times for digital circuits,plus the small size and increased sensitivity ofanalog circuitry, presents an opportunity forElectra-Magnetic Interference (EMI) when COTScomponents are used in avionics applications.The increased functionality that can be fitonto a module presents the potential for a highdensity of circuitry in close proximity.Precautions must be taken to prevent analog anddigital circuitry from interfering with each other.Moreover, the newer, higher-speed circuitry maypresent EM1 levels that exceed the levels thatlegacy avionics were designed for. Thus,precautions must be taken to prevent the avionicsupgrade and the legacy avionics from interferingwith each other.Because of the rapid advances incommercial computer technology, COTS-basedprocessors should be the most cost-effective partof an avionics upgrade. It is reasonable, therefore,to examine which traditionally analog avionicsfunctions can be handled inexpensively by theseprocessors. Digital radios are becoming availablein the commercial arena, so it makes sense o pushdigital technology as far as reasonable in avionicsupgrades. Digitizing at the aperture couldeliminate costly waveguides and high-maintenance analog hardware, replacing it with

digital networks and signal processing. This

8/8/2019 EN-014-04

10/16

4-10

approach to upgrades may reduce cost, weight,and volume, while enhancing maintainability.Quality Assurance - Quality assurance(QA) objectives of military systems must bemaintained when applying commercial productQA standards established by the InternationalStandard Organization (ISO). It has beendemonstrated that streamlined COTS processesand procedures can reduce Life Cycle Costs andalso meet the needs of military applications whenapplied appropriately.Operational Requirements - Applicablethermal, random vibration, pressure/altitude,electromagne tic and other military operationalrequirements must continue to be satisfied whenCOTS elements are introduced in weaponsystems. A number of COTS products arepresently qualified for many legacy aircraftsavionics operating environments. The associatedCOTS suppliers have implemented engineeringoptions for meeting different operationalrequirements.A technology bridge must becontinuously maintained into the design ofincremental upgrades, so that parts obsolescenceand low cost COTS insertion can become facts of

life to legacy avionics systems.Obsolescence and Rapid Design Changes

The life cycle of microcircuit technologyis getting shorter and shorter. In the 60s themarket availability life expectancy was 20-25years. The market availability life expectancy ofmicrocircuits from the 70s was reduced to 15-20years. This was further reduced in the 80s to lo-15 years. Today the market availability lifeexpectancy of new microcircuits is 7-10 years andthis includes introduction and phase-out. Inreality, for some of the high-performancemicrocircuits like processors and memories, themarket availability life expectancy is less than 5years. This presents a problem for militarysystems, which are generally designed to operatefor a life span of 20 to 30 years.It is not uncommon for military avionicsto be obsolete before it is deployed. Many modernsystems, including the F-22 and C-17, haveundergone major avionics upgrades beforeachieving operational status. Between the changes

in requirements due to the global and nationalpolitical climates and the rapid evolution oftechnologies applicable to avionics, the customarchitectures created to solve the originalproblems become rapidly outdated and veryexpensive to change.Some of the major cost factors for obsolescenceand design changes are described be low:l The current rate of technology turnover in thedigital circuit market is staggering. The meantime to obsolescence for all digital parts isaround 24 months, and processors arebecoming obsolete in less than 18 months. Formilitary avionics, this situation iscompounded by the fact that there are adecreasing number of silicon foundriesproducing military grade parts, and thosefoundries are producing fewer military parts.It is simply not cost effective for thesecompanies to tie up valuable productioncapability for a minority market segment.Military parts do not drive the market fordigital circuits, and the military supplier mustlearn to use commercial digital parts to deliverfuture avionics systems and avionics systemupgrades.Digital components are not the only systemcomponents that are affected by technologychange. Software interface standards, busstandards, protocol standards, network

standards, and others all change at differentrates. Each of these system components arecost risk factors for new or upgrade avionicssystems.l The non-recurring engineering (NRE)required to capture the existing systeminterfacing requirements is non-trivial. Thesystem documentation does not contain thewisdom of the originaldesigners/implementors, and many timesthese people are no longer available . Further,the documents do not reflect the current state

of the system -- or are missing the fine detailsto make the system really work. Accuratespecifications of the system interfacing andfunction are required to upgrade avionicssystems.l Beyond the problems of understanding thesystem interfaces and function, the upgradeavionics system must pass several re-certification levels. Flight qualification,system acceptance, and a number of othertests are required before the upgraded systemcan be deployed. The cost of re-certification

8/8/2019 EN-014-04

11/16

4-l 1

of avionics systems can be a major costdriver .l One of the major operations and support costis the logistics to track and store spare partsfor aircraft systems. If lifetime buys are

considered, several cost factors must becontemplated. First, the sunk cost of thelifetime buy. Second, the additional logisticscost for tracking and storing more parts.Third, the cost risk of those parts becomingobsolete before the parts are used. Finally, thecost of monitoring the obsolescence status ofparts within a design. All these cost factorsmust be considered for the lifetime buyoption.Parts obsolescence is also an issue formilitary components but not nearly as severe as itis in the highly-competitive COTS environment(see Figure 5). Provisions for the management ofobsolescence need to be in place from thebeginning of the design process. Obsolescenceconsiderations must be part of any trade-offstudies performed in order to select candidateCOTS technologies for use in a military systemwith an extended operational life. One way tomanage obsolescence involves drastic changes inthe way the military acquires systems. It entailshaving a number of short development and

procurement cycles throughout the life of the

system, where upgrades are incorporated into thesystem. These short development cycles wouldinclude the required testing to recertify theintegrity of the system. If done right, thisapproach can eliminate the need for many of thelogistic support requirements. If the equipment ishighly reliable and low cost, only a small numberof spare parts will be required to sustain thesystem in operation. Upgrades could be availablebefore any failures occur, eliminating therequirements for a complex log istics supportstructure.

A planned obsolescence approach isrequired for systems using technologies with ahigh rate of change. One approach to address thistechnology change is to develop an overall systemarchitecture that uses well defined (and widelyused interfaces) for those items that are subject torapid obsolescence. Widely used interfaces willtend to have support strategies to get to the nextstandard (e.g., interface chips that bridge the oldinterface to the new one). Less popular (or evencustom) solu tions can be used for systemarchitecture components that do not changerapidly (i.e., antennas or radio frequencyamplifiers).

From this baseline upgrade open

Typical MILDevelopment/ProductIon

COTSDevelopment/Production

HIGH

MED

1 2 3 4 5 6 7 8 9 10

PotentialRedesignDue toObsoleteComponents orP31 Improvements

1 2 3 4 5 6 7 a 9 10YEKS

Figure 5 Contrasting Program Costs

8/8/2019 EN-014-04

12/16

4-12

architecture, portions of the architecture can beinserted incrementally until the system achieves afully upgraded state. Then as components of thatarchitecture become obsolete, they can be easilyreplaced with newer generation parts. Part of thisapproach is the establishment of an obsolescence-management contractor-based sub-IPT early in theprogram. The IPT would be in charge of:Performing detailed obsolescence surveysof potential suppliersMaintaining current survey results on fileIdentifying single point-of-contact forobsolescence notification in each of thesub-contractors involved in the programMonitoring and maintaining life-cycleratings for all the active technologiesbeing used in the programRating of technologies longevity byfamiliesPerforming program baseline reviewsEstablishing and maintaining a masterparts file repositoryRecording and maintaining a lessons-learned file.A design capture process using model-in-the-loop techniques also helps solve problemswith obsolescence of commercial technology-based upgrades. The design capture processresults in a detailed executable specification thatcan be verified using system acceptance tests andother real-world inputs. This executablespecification can then be implemented in variousways, including a VHDL-derived replacementpart, software emulation, or a modern-technologysoftware simulation. When the first

implementation becomes obsolete, due tounavailable parts or for other reasons, a newinstantiation of the system can be derived from theexecutable specification at a substantially reducedcost. This new replacement may or may not usethe same implementation choice as the firstupgrade, but since the executable specificationalready exists, little or no time needs to be spentderiving requirements or defining test procedures.By using executable specifications (e.g.,VHDL) to describe digital components,

replacement parts that meet the functional

specifications can be achieved. Commercialindustry routinely uses VHDL to fabricate orprogram gate arrays and other programmablelogic families (e.g., FPGAs, PALS, GALS, etc.)via commercially available synthesis tools. Whilenot perfect, this process can be used to describelarge levels of functionality (e.g., processor andI/O boards). Other interfaces (e.g., physical,electrical, . . .) can be achieved through the use ofcustom packages or adapters. This is approach ispossible because of the rapid size and form-factorreductions being seen in the programmable logicarena.Open System Implementation

The primary system implicationsassociated with the use of COTS/BestCommercial Practices occur at the avionicsarchitecture level. A major concern is that, takento the extreme, reduced contractor guidance couldresult in a hodgepodge of custom boxes,modules, display, etc. which will create anintegration and maintenance nightmare, destroycompetitive procurements (competitors cannotdetermine how to build compatible hardware andsoftware) and make common, interoperableavionics impossible. Obviously, measures mustbe taken to avoid this situation.One measure that has the potential ofreducing some of these problems is the use of anOpen System Architecture (OSA) for future andcurrent systems. Definitive guidelines of howOSAs will be employed are currently underreview by a DOD task force. The OSA approachis aimed at ensuring maximum competition andcommon avionics, through a readily available setof system specifications that provide adequateinformation to build interface-compatiblehardware and software.The key is to define an OSA that willpermit an incremental upgrade strategy utilizingmodular, scaleable components which can beinserted into the diverse set of fielded legacyarchitectures. This architecture will provide aframework that can be installed in affordableincrements into legacy aircraft while retainingcoherent system functionality. The ability toincrementally upgrade a system is important, notonly due to budget realities, but also because it istypically impractical to complete a major system

insertion/replacement within a single normal

8/8/2019 EN-014-04

13/16

maintenance cycle. Aircraft downtime for kitinsertion, functional validation and training mustbe taken into consideration for reducing time-to-field in both engineering and manufacturingdevelopment and production phases of anupgrade.The incremental OSA approach graduallymigrates the legacy system to a new coherent,easier-to-upgrade architecture that avoids thecurrent patching problem. These incrementalupgrades can be accomplished in less time and atlower cost than would otherwise be imposed bytraditional upgrade methods. Closely related to theinstallation problems (Group A) described earlierare functional interface issues. If incremental andcontinuous upgrades are to be successful, welldefined functional interfaces must be establishedbetween the portion that is being upgraded and theportion that is not (see Figure 6). The purpose ofthese interfaces is to establish boundaries so thatthe upgraded portion may not affect the

4-13

difficult to apply to existing systems on anincremental basis. Open systems, properlyexecuted, promote competition and provide valueto the customer at lower cost. Distributed opensystems are an alternative that bears study, alongwith domain-specific architectures for displaysand display processing, data and satellitecommunication links, and mission processingupgrades. Specific aspects of an upgradearchitecture, such as software re-hosting, wouldbe applied system wide. The issue is the cost ofchanges to the existing system, including test andmaintenance equipment, procedures, anddocumentation, versus the benefit of a newintegrated architecture.Open interfaces that satisfy widely-usedstandards are essential to implement incrementalupgrades. One upgrade can then benefit from theinvestment and knowledge base of the previousone, such that we do not waste money solving thesame system problem over and over aga in.

Figure 6 Upgraded Avionics Integrated Through Well Defined Open Interfacesfunctionality of the portion that is not beingmodified. If this is not done, a ripple effect islikely to occur where changes affect parts of thesystem that dont need to be changed. Not onlydoes this unnecessarily increase the life-cyclecost, it also leads to the unwanted piecemealdesigns described above. While new weaponsystem developments are emphasizing openintegrated architectures, these concepts may be

Previous attempts at common module and openinterface programs have failed to provide for thefuture insertion of rapidly evolving digitaltechnology, or have selected interfaces whoseusage is limited almost entirely to militarysystems.Avionics upgrades generally involve

changes to both hardware and software, usually to

8/8/2019 EN-014-04

14/16

4-14

accommodate new functional or performancerequirements, but occasionally to reduce cost ofownership. We expect that affordable upgradeswill become even more of a challenge as thenumber of software lines of code in our fieldedsystems increases dramatically during the nextdecade. Piecemeal, single-platform upgrades arehere to stay as long as the acquisition process isgoverned by individual program offices andfunding horizons are short term. Incrementalupgrades are desirable over piecemeal upgradesgiven the reality of the current budgeting process.Present approaches do not institutionalize therelevant methodologies and architecturalconcepts, but rather force them to be relearnedeach time by individual contractor teams. Theresult is a much longer and more expensiveupgrade activity, every time.

SummaryIndustry has embarked on a product-lineimplementation strategy. The major WeaponSystem/Aircraft Manufacturers are in the process

of defining an open system avionics architecturebased on commercial technologies and processeswhich will be applied across the product lineswhich they manufacture. Much of this activity isfocused on aging aircraft avionics issues to extendthe service life of current aircraft coupled withdeclining military budgets and a dwindlingsupplier base which challenges the effectivenessof todays front line weapon systems. Thisapproach has been developed by working withnumerous government agencies and across theservices as well as internal company funding toproduce a focused strategy for maintaining theeffectiveness of legacy systems while charting afuture business strategy similar to largecommercially oriented corporations.

Effective use of advanced technologiesderived through Science and Technologyprograms can only be accomplished through acontinuous Demonstration and EvaluationProgram. This program must be agovernment/industry partnership aimed at thetransition of needed advanced technologies fromthe research environment to the warfighter. (SeeFigure 7)

Transitional PrototweMulti-phase, Multi-Year Programswith WSCEmphasis S&T

Figure 7 S&T Transition Methodology

8/8/2019 EN-014-04

15/16

4-1s

References1. Low Cost Avionics Through Best CommercialPractices, Ott 1995, Advanced ArchitectureGroup, Avionics Directorate, Wright Laboratory.2. Impact of COTS on Military AvionicsArchitectures by J. Carbonell and J. Ostgaard,AGARD Conference 581, July 1997, (AdvancedArchitectures for Aerospace Mission Systems).3. The Legacy Systems Challenge, DefenseNews, Gen Laurence A. Skantz, November 1998.4. Scientific Advisory Board Summer Study( 1997) on Aging Systems

5. Supplemental Cooling for Legacy AircraftAvionics, Stephen Benning and John Ostgaard,November 1999, Digital Avionics SystemsConference.

8/8/2019 EN-014-04

16/16