Artemis-2011-1 decision and platform support for model‐based eVolutionary development of Embedded systems Date of preparation

Section 4 - Market innovation and market impact

4.1 Impact

The DECISIVE project facilitates the transition from a vertically structured market to a horizontally structured market by focusing on software engineering of complex embedded systems as a crosscutting system discipline that transverses many traditional product and service segments. DECISIVE does not focus on any single application domain; instead our results will have high impact on all application domains where embedded software is a driver for growth and increasing competitiveness. In these segments, complexity and size of software makes an evolutionary approach to software development paramount to master increasing demands on cost-efficient development, cost-quality balancing and time-to-market.

The DECISIVE project contributes to the Artemis targets in the following way:

The project addresses the reduction of the cost of system design by 10% by providing a model and tools that bridge the gap from abstract design models to advanced hardware platforms with e.g. multi-/many-core, DSPs and GPUs, while still allowing the performance characteristics from such hardware to be modelled and analysed at the abstract model level.

The project addresses the reduction of the development cycles by 25% by improving the development throughput and productivity by automating collection of system properties to support evolvability and decision making.

The project addresses the target of managing an increase of complexity of a factor 3 with an effort reduction by 10% by providing methods and tools that improve ever increasing product management and upgrade requirements. This will be possible thanks to educated and evolved decision support with respect to product management and system architecture.

As a tangible result of the investigations and developments in the DECISIVE project, companies expect to achieve a higher productivity throughout the workflow of IC development. Key enabler is the more flexible modelling and the extended validation features on system architecture level. This will shorten the design cycle of technical demands in the product generations to come. It will bring complex and expensive development closer to “first time right” results and accordingly improve time-to-market significantly. Early feedback on the verification coverage of new designs based on evolutionary capabilities of models and a homogeneous concept for the whole system will make this possible. The environment of, e.g., integrated circuits can be included in a simulation to a larger extent, which supports early detection of non-compliance or interoperability issues. Contact-less identification chips for instance have by default a very complex power management. Only thorough estimation over all elements continuously updated with evolutionary data will help to reach higher efficiency, thus better performance and save cost.

To quantify the above statements exemplified for the semiconductors industry the following rule of thumb calculation can be made. A new chip design (new platform) in contemporary technologies will have development cost in the range of 50+ M€¹², every re-design cycle requires a Non-Recurring Engineering (NRE) of about 500k€, a simple correction cycle about 200k€, cost for test and verification after each cycle will be in the same magnitude. The time needed for another redesign, including production and test and verification afterwards will delay a new product for at least 9 months if not for a whole year. Furthermore, the commercial loss due to late delivery can be significantly higher.

Contribution to the degree of application innovation

Embedded systems are increasingly important within Europe. The DECISIVE project aims to improve the competitiveness of the European industry through the improvement of the development of embedded systems in many application areas, including (but not limited to) the areas represented by our industrial partners: healthcare, automotive, rail, aerospace, telecom and manufacturing.

As a tangible result of the investigations and developments in the DECISIVE project companies expect to achieve a higher productivity throughout the workflow of embedded systems development. Key enabler is the more flexible modelling and the extended validation features on system architecture level. It will bring complex and expensive development closer to “first time right” results and accordingly improve time-to-market significantly. Early feedback on the verification coverage of new designs and a homogeneous concept for the whole system will make this possible. The environment of the embedded system can be included in the simulation to a larger extend, which supports early detection of non-compliance or interoperability issues.

Philips: The results of Decisive will be used to produce high reliable medical embedded systems for a lower development and maintenance cost. This reduces the cost-of-ownership of such systems. The first targeted systems will be used for the growing field of image guided intervention. Philips will raise its share in the interventional imaging market by offering integrated solutions. Through Decisive Philips will also be able to create a position in the expanding and profitable delivery system and therapy business areas. On basis of successful IGIT market propositions, Philips expects to generate €500M extra annual sales in 5 years

General the healthcare area:

Global economic growth: increased spending on health related services, access to healthcare for a larger number of people and increased awareness of available healthcare options

Dramatic changes in demographics; aging population:

15. 
    By 2045 more people will be over 60 than under 15 years, rising from 600 million to 2 billion.
    
16. 
    Rise in number of patients with age-specific, chronic and degenerative diseases (cardiac, cancer, diabetes, Alzheimer’s, Parkinson’s). The number of US patients with a chronic illness grows from 118 million in 1995 to 157 million in 2020. For Europe, a few key numbers are (Frost & Sullivan 2005):
    
17. 
    neurodegenerative diseases: 3,600,000 people affected with Alzheimer
    
18. 
    cardiovascular disease: 460,000 deaths of stroke
    
19. 
    oncology: 240,000 deaths for breast cancer
    

Efficiency and effectiveness of healthcare: need to further improve hospital work flow efficiency, integration of diagnosis and treatment. E.g. the average length of stay for acute care has fallen in nearly all OECD countries - from 9 days in 1990 to 6 days in 2005

Skyrocketing healthcare costs: global health care spending expected to grow from 9% of world wide gross domestic product (GDP) in 2006 to 15% by 2015

The global market for medical imaging (diagnostic and interventional imaging) is estimated to be $20B (2007 TriMark study). The European market is about a quarter of this total and the US market almost half. The medical imaging market records solid growth percentages. Depending on the modality, the average compound annual growth rate (CAGR) is about 4% (for interventional imaging this is 8%). There a few specific areas where growth is markedly higher than average.

Image-based software applications that support intervention processes in healthcare have these growth opportunities:

The European market for 3D/4D imaging software has a CAGR of 14% from 2004-2014

The global CDSS market grows from €159M to €289M during 2006-2012 (Frost & Sullivan)

OCE: Océ will use the results of the DECISIVE project to in the development of new products. For Océ the main benefit of model driven design will be a shorter development time of new machines and improved machine capabilities.

Models increase understanding across developers of different disciplines, enable early validation, and make it possible to assure that well-tested control software exists at the very moment the mechanical machine prototypes (lab models) are ready. Energy consumption during usage and standby can be reduced significantly when intricate trade-offs between several system aspects have been modelled and evaluated extensively. Furthermore, the use of virtual prototypes significantly decreases the amount of waste (physical prototypes, piles of test paper for printer stress testing) during product development.

Earlier time-to-market increases the amount of machines that can be sold. Even a small decrease in time-to-market is very significant for projects with a lead time of a few years (and a hundred developers). This has a strong positive effect on the competitiveness of Océ.

Almende: The results of DECISIVE will be used to develop of self-organized critical agent-based solutions that sustain and improve the coordination of communication and collaboration across evolving networks of humans and ICT systems. These solutions will be applied to several application domains, including healthcare, logistics and crisis management.

NXP Semiconductors: As a tangible result of the investigations and developments in the DECISIVE project NXP expects to achieve a higher productivity throughout the workflow of IC development. Key enabler is the more flexible modeling and the extended validation features on system architecture level. This will not only to improve the mixed signal flow for the next range of standard mixed signal designs, but also shorten the design cycle of these tough technical demands in the chip generations to come. It will bring complex and expensive development closer to “first time right” results and accordingly improve time-to-market significantly. Early feedback on the verification coverage of new designs and a homogeneous concept for the whole system will make this possible. The environment of the IC can be included in the simulation to a larger extend, which supports early detection of non-compliance or interoperability issues. Contact-less identification chips for instance have by default a very complex power management. Only thorough estimation over all elements will help to reach higher efficiency, thus better performance and save cost.

Valeo: Validate on real automotive developments an Artisan Studio prototype which addresses safety analyses (based on safety standard norms) in the early stages of system design.

Support the DECISIVE methodologies (from the specification to simulation and analyses) by developing uses cases in Artisan Studio

To quantify the above statements for the semiconductors industry the following calculation can be made. A new chip design (new platform) in contemporary technologies will have development cost in the range of 50+ Million EU, every re-design cycle requires a NRE of about 500k EU, a simple correction cycle about 200k EU, cost for test and verification after each cycle will be in the same range. The time needed for another redesign, including production and test and verification afterwards will delay a new product for at least 9 months if not for a whole year. Despite the additional cost, the commercial loss according to late time to market can be much higher. Production time and cost can hardly be cut down, but the amount of money to be saved by more effective validation and verification, faster design flow and a reduced need for correction cycles, would certainly be significant. Faster time to market will not only allow a bigger share of the market and better prices, but also strengthen the position of European companies against strong competition from Asia.