SELEX sensors and airborne systems
SELEX sensors and airborne systems
Keywords: Defence sector, Aircraft components, Radar
Devices intended for use in defence aircraft, vessels and vehicles are rarely off the shelf. Instead, developing devices like sensors and radar for today's fighting vehicles has traditionally required starting from scratch to develop bespoke, tailored systems (Plate 1).
Plate 1A comprehensive self protection suite for the Eurofighter Typhoon aircraft
However, as the industry continues to require ever-quicker development,debugging and delivery for systems and devices that provide crucial tactical advantage – leading manufacturers too are looking to optimise tactics in a battle for dominance. Continually ensuring faster and more efficient delivery of sensors and radar devices for avionics including safety-involved and safety-critical systems has meant looking ahead of the curve to be first in re-evaluating common practice.
As a leading provider of airborne radar systems, electronic warfare systems,military lasers and electro optical systems – SELEX Sensors and Airborne Systems (S&AS) decided to take on an uncompromising, forward-looking strategy to ensure they stay one step ahead of increasingly demanding projects embarked on with leading primes like BAE Systems.
Previous assumptions about project management and product development were re-examined and solutions to common development overhead and project delays were taken apart to find alternatives. Assumptions of fixed costs in terms of time and money were scrutinised to find new efficiencies. (Plate 2)
Plate 2Covert reconnaissance of illegal activities
Among the key realisations was that SELEX S&AS as an enterprise was spending substantial portions of product development time and resources creating tools and software components that did not have a direct impact on the quality or functionality of the finished product. Russell Shoulder, Engineering Functional Manager led the effort to review processes and points out, “We were creating tools and commodities, such as the Real Time Operating Systems(RTOS) in-house. Our engineers were also creating bespoke tools and hardware support for individual projects.”
The lack of standardisation of tools and software components meant that every project start-up required identifying, justifying and costing solutions along with addressing the inevitable training needs for developers coming onto the project. Certification for ARINC 653, the safety-critical standard for avionics,was also complicated by relying on a patchwork of ad hoc fixes to hardware/software interfaces and tailored, DIY RTOS all authored by SELEX S&AS engineers to create devices from the ground up.
Shoulder led the effort to implement Device Software Optimisation (DSO)throughout SELEX S&AS, “We decided to standardise wherever possible on VxWorks as a workbench across the enterprise. We have used other operating systems, but VxWorks is the best operating system for military applications. It's a ruggedised product.”
Certification has also been facilitated by re-using common tools and diagnostics across projects - proven solutions were needed to replace impromptu patches. As Shoulder explains, “In the past, we've used or own operating systems because we had no choice. Certification was very expensive and difficult. It distracted engineers from their core skills: real- time software engineering, not operating system development. Now we have commercial products that are approved and accepted.”
DSO also helped SELEX S&AS to implement coherent processes for working with open source. While Linux is known for frequently forking into incompatible strains – SELEX S&AS chose commercial-grade Linux from Wind River to ensure that all open-source solutions started from a reliable, debugged foundation. The push for concurrent engineering schemes and open architectures for high-priority, ongoing sustainment projects with large numbers of sub-contractors, is leading the defence device industry to embrace platform-agnostic standards, including DSO and open source (Plate 3).
Plate 3LINAPS in action during the second Gulf War
While open source is usually regarded as being “free” Shoulder explains the rationale behind choosing commercial- grade: “Open source is the future for any large sector. Reliability is the main reason we would select it and its widespread use is an advantage. We require a commercial distribution because updates are configured and tested before we get them. We require a formally configured product. It also makes certification easier because it's an industry standard.”
These bigger, ongoing projects also require a higher degree of flexibility from all the sub-contractors. Designs must be able to adapt to sudden changes in specifications. Optimising device software across an enterprise has allowed this versatility. As technical requirements change suddenly, COTS device software components are interchanged without time impact.
“Workbench tools, diagnostics and RTOS – the foundation of device software, are now COTS components shared across all SELEX S&AS projects. The DSO strategy also ensures that resources are scaled according to project demand. In the past, unusually difficult projects meant a strain on resources as developers put in long hours to find workable solutions and came in early to debug and validate those solutions. Now SELEX S&AS can call in vendor support to outsource the effort and cost,” says Shoulder.
Shoulder explains the reason for doing away with one-off tools and solutions,“If you're training people in the use of tools, it's very expensive. It's an investment. We're trying to cut out extra human intervention. We want our engineers to work on a higher level: modelling; architecture; not writing bespoke tools. By definition, an in- house RTOS has to be tailored to every application. Standardising offers critical savings because you don't have to modify or revalidate.”
Shoulder points out that steady acceleration is one thing the industry can take for granted. Appropriating a comprehensive development strategy is key to leveraging existing resources. Keeping pace is not a sustainable goal, SELEX S&AS chose to take the lead and maintain it. Shoulder reflects on the changing needs of today's customer, “The product development cycle will only improve. Ten years ago a radar took from ten to fifteen years to develop. Now it's two to three years for the whole industry. The whole environment is speeding up with tools and productivity. For SELEX S&AS, creating an operating system with board support packages and configuring it with target hardware used to take four months. Now, with DSO and vendor support, it takes two to three weeks.”
