26th Annual INCOSE International Symposium (IS 2016) Edinburgh, Scotland, UK, July 18-21, 2016
Learning Systems Engineering Lessons from an Electric Vehicle Development Gerhard Swart Director, Alphadot (Pty) Ltd 31 Second Ave, Melkbosstrand, Cape Town, South Africa Telephone: +27 (83) 291 5148 [email protected]
Copyright © 2016 by Gerhard Swart. Published and used by INCOSE with permission1.
Abstract. The Joule Electric Vehicle (EV) was developed “from a clean sheet” by Optimal Energy, a South African start-up company established in 2005. In the seven years to its liquidation in 2012, the company grew to 108 people, built up significant international partnerships, developed substantial in-house technology and performed three development iterations culminating in four market-testable prototypes. Developing a vehicle from scratch requires a different approach to the traditional automotive development processes, which are optimised for evolutionary vehicle improvements and cost reduction. Developing a new EV in a new organisation without legacy processes or manufacturing infrastructure presents additional challenges but also provides opportunity to work “top-down”, focussing on the real “mission requirements” instead of being trapped by legacy practices, design concepts and infrastructure. Systems engineering “best practices” may be applied to a larger extent and the concept of passenger mobility can be optimised for electric propulsion instead of accepting the traditions of fossil-fuel power. This paper provides highlights of the unique challenges and opportunities related to EVs, and presents the tailored development process that was established for the Joule EV.
Introduction The author has in a previous paper introduced Optimal Energy as a start-up company with the purpose of “establishing and leading the Electric Vehicle industry in South Africa and expanding globally” (Swart, 2015). The paper reports on the funding, organisational highlights, commercialisation requirements, development process and subsequent difficulties of the organisation. After briefly re-introducing the Joule EV itself, the author will now present a typical automotive development process and highlight several challenges that indicate the need for a different process for the Joule EV development. Finally the author will show how the automotive process was merged with a tailored Systems Engineering process to become the development process for the Joule.
Originally presented at the INCOSE South Africa 2015 Conference under the title “Learning Systems Engineering Lessons from the Joule EV Development”, but re-written to reduce its regional focus.
The initial EV concept developed by the company was minimalistic and not particularly attractive, but as the real customer needs and desires were understood the Joule emerged as an attractive mid-sized passenger car, with competitive performance and features (see Figure 1 below). As a full 5-seater C-segment vehicle with a top speed of 135km/h and 0-60km/h acceleration of less than 5s, the Joule was competitive with its petrol and diesel cousins. It was designed to achieve an NCAP 5-star safety rating and came with a luxurious internet-ready telematics suite. It was well received at various international automotive shows and after Car Magazine’s test team drove several prototypes in 2011 they featured it in the April 2011 edition, concluding: “It’s good. Very good in fact.” (Oosthuizen, 2011).
Figure 1. The Joule EV Exterior and Interior A detailed description of the Joule and its technologies is beyond the scope of this paper, but Table 1 below provides some its key characteristics. Table 1. Joule key characteristics Top Speed: 135 km/h Battery: 32kWh, 350V, 200 000km life Acceleration 0-60 km/h: 5s Range per charge: 230km (NEDC cycle.) Vehicle class: C-segment MPV with 5 seats Recharge time: 1h with off-board charger, and large luggage compartment 8h using household 220v Braking: regenerative and discs with ABS, User Interfaces: Integrated info-telematics, 100-0 km/h in 3s internet connectivity and ability to load Apps Target Price (excl. battery which follows a Electric motor: STM type, 70kW peak rental model) : €26 000 (before consideration of subsidies) Safety rating: airbags, NCAP 5-star Options: three luxury levels, PV on roof, small or large battery Four third-generation roadworthy prototype vehicles (called PEV) were built (Figure 2), incorporating the key Joule features and technologies, for the purpose of validating requirements in the target market and evaluating various aesthetic and technical concepts. These vehicles completed 38 000km of testing before the company was closed in 2012.
Figure 2. The Four Roadworthy Joule PEV Prototypes The PEV prototypes were the forerunner to the vehicle industrialisation phase, which would re-engineer the vehicle and its parts to reduce cost and establish manufacturing capacity to produce 50 000 per year, the estimated volume required to achieve profitability. An extensive multi-national team was built to meet this development, marketing and manufacturing challenge.
Legacy Automotive Development Approach Although every automotive manufacturer produces unique vehicles that embody their specific brand values, each vehicle having emerged from differing development processes and having divergent manufacturing, costing and marketing strategies, the author will make some generalisations for the sake of comparison. Statements concerning the existing automotive industry are thus not based on conclusive research but should rather be seen as generalisations and considered opinion that have developed through several years of direct interaction with the many automotive engineers and organisations. Although South Africa does have a number of local assembly plants for conventional vehicles (most of which also serve the export market), it should be noted that no complete vehicle development facility existed locally prior to the founding of Optimal Energy. The key objective for any business is making a profit for the shareholders. Global automakers are no different and achieve this mostly through after-sales income. The vehicle itself is actually quite expensive to manufacture and the margins are low. This is especially true considering that a car may comprise of more than 1800 parts, each requiring a factory to make it. This creates great technical and organisational complexity, where the vehicle assembly plant operated by the OEM is simply the tip of a huge supply-chain iceberg. In small and medium sized vehicles (the so-called A, B and C segments), where there is fierce market competition and limited brand premium paid by the customer, it is particularly vital to have high-volume manufacture to achieve economies of scale. It has thus become a fundamental practice for automakers to re-use as many components as possible between their various models, even sharing parts with their competitors. What is unique between successive vehicle models is most often only the aesthetic styling and minor variations introduced to create a unique experience for the customer. It is not uncommon for successive models to have exactly the same steel body, suspension and engine, but to have different lights, interior trimming and different exterior plastic trimming.
As a result, the typical process for developing a new vehicle model is an evolutionary one, starting with what is available. Even a radically new model would normally retain two of the three major cost/risk items: platform (chassis, suspension and interior); drivetrain (engine and gearbox); or assembly plant. The development processes of the major automotive companies are kept confidential, but Magna Steyr in Austria does contract development and manufacture of entire vehicles for OEMs, and have published their process. Figure 3 shows their vehicle-level development process from concept up to Start of Production (SoP).
Figure 3. An example Automotive Product Development Process (Magna Steyr, 2015) What feeds into this development process is the high-level vehicle requirement (including target market, new styling, constraints on re-use of a prior model “carry-over parts” and assembly plant details) and newly available part technologies that may provide an advantage. In defining the Joule EV development process it was necessary to question whether this process, which is largely evolutionary, was suitable for a new company with no legacy products, parts or production infrastructure. However, if a pure top-down Systems Engineering approach was followed, it would still be important to maximise “off-the-shelf” parts to achieve the product cost targets. In addition it would be necessary to retain elements of automotive terminology and processes to allow engagement with the existing automotive industry. But first one had to decide whether an EV was simply the electrification of a “normal” car, or whether a clean-sheet approach was required to achieve the design goals.
Are Electric Vehicles Really Different? Although some of the first automobiles invented were actually electric, the abundance of fossil fuel and the Henry Ford legacy has for a century shaped cars to be mechanical devices with electronics added for convenience and safety. The hot, noisy engine is the centrepiece that dominates the vehicle layout because of its size, weight and noise. Through the years electronic features such as electric windows, air bags and electronic engine controls were progressively added. Loureiro found that for modern vehicles rich in electronics and software, “an interdisciplinary, collaborative approach to derive, evolve and verify a life-cycle balanced
system can deliver better results that meets customer expectations and public acceptability. This approach is systems engineering.” (Loureiro, 1999). Two additional reasons to consider an alternative development process are the disruptive nature of Electric Vehicles, and the specific South African context, where a small company had to penetrate a market against the established automotive industry.
What the Customer Wants It is easy for engineers to think that they are a representative sample of their product’s market. It was a hard lesson for the Joule team to learn that “the voice of the customer” must actually be allowed to speak, even against better technical judgement, to have the final say in a product. Nowhere is this more important than in the car industry where vehicles are purchased by housewives, doctors, scientists, shopkeepers and even grandmothers, from all walks of life, most with little or no technical insight. The final purchase decision is normally an emotional one, dominated by subjective elements such as the “feel” of the vehicle, visual attractiveness and even the sounds and smells. Functionality is important, but in the end, the buyer desires the product and if it is affordable (or perhaps even if it is not), will purchase. It is, however, the way a vehicle looks that is arguably the most important instrument to awaken the purchasing desire in the potential customer. The emotional appeal of a vehicle’s appearance is overlooked by many engineers and led many earlier EV attempts to their grave. Converting an existing fossil fuel vehicle into electric simply would not break the pre-conceived impressions of the original petrol/diesel vehicle. Through the help of Keith Helfet, the renowned Jaguar designer, Joule was conceived with a fresh and unique look, and judging from the global media and public response, awoke the desire that was needed. In establishing whether an EV deserves a fresh engineering approach, let us also interrogate three aspects that may or may not be unique in the eyes of the customer: the purchase price of the vehicle, its mechanical architecture and its electrical architecture.
Pricing an Electric Vehicle Introducing a totally new EV like the Joule at a competitive price, in the face of established competition, is difficult. The new technologies incorporated into an EV, particularly the Battery System, Power Electronics and Drive Motor, had not yet benefited from high-volume global production. Although one could save around €6 000 from the cost of a typical sedan by leaving out the engine, the additional cost of these EV components, even when making 50 000 per year (the planned Joule production rate), adds back more than €14 000. For most customers this €8 000 premium would prohibit them from switching to an EV. This dilemma has held back EV development in the automotive industry, to the point that in 2008 Nancy Gioia, then Ford Motor Corporation’s Director of Global Electrification, said: “We now believe that Electric Vehicles are the way forward, we just don’t know how to make money from it.” 2 If however, a new approach to the business model is followed, it suddenly does make sense: A vehicle owner travelling, 30 000 km per year to work and back, who switched to an EV, would pay around €550 for electricity instead of €2 350 for fuel, saving an estimated €1 800 per year and have lower vehicle servicing costs. If this saving was used to purchase the battery over its lifetime (i.e. leasing it) then the vehicle price could be reduced to a competitive level. 2
Heard by the author at the 2008 SAE conference where Nancy Gioia was a speaker.
Establishing the relationship between the performance that the customer desires (e.g. vehicle range, acceleration and top speed) and the commensurate price they would be willing to pay is non-trivial. The technical implications of the customer’s desires have to be determined and filtered through engineering trade-offs and fed into the business model before establishing them as formal target requirements. Ultimately though, the asking price for the Joule would be established through a combination of market research, “market clinics” with potential customers and benchmarking against competitors. Out of the mix, a vehicle cost/performance combination was found that was incorporate into the User Requirement.
Vehicle Mechanical Architecture As alluded to earlier, the mechanical layout and structure of a conventional vehicle is largely influenced by the large, heavy, noisy device which we call the engine. Its need for frequent maintenance drives the requirement for a hood and engine compartment, whilst its bulk and mass impact the crash-safety design and driveability of the vehicle. EVs on the other hand, require very little maintenance, the electric motor is small and cool, whilst instead the battery is large and heavy. Trying to package the EV parts into a conventional vehicle body is thus clearly sub-optimal. Some automotive companies initially failed to recognise this, resulting in EV offerings with severely compromised handling through the increased vehicle mass and luggage compartments half-full of batteries. The Joule mechanical layout centred on the occupants, but the battery was the second most important consideration to make it “Born Electric” 3 and give the customer a true EV experience instead of a compromised one. The first prototype, PT1, (shown in Figure 4) was used primarily to establish an understanding of EV technologies and the integration issues such as the battery location.
Figure 4. Joule PT1while testing The battery compartment is seen under the floor between the front and rear wheels. This keeps the vehicle centre-of-gravity low, enhancing vehicle handling and providing access to the batteries for battery swapping. 4 The under-floor battery compartment is viewed from 3
Today this is a registered trademark of BMW. Battery swapping was seen as a better alternative than “fast-charging” for quickly replenishing the vehicle’s energy store. Tesla Motors later also later decided to follow this route (Capgemeni Consulting, 2014). 4
underneath in Figure 5 and forms a strong structural element providing significant vehicle stiffness when the closure panels are fitted. Without the need for a fuel tank and routing of an exhaust pipe, the Joule could have a totally flat floor allowing great flexibility in seating layout and allowing the same platform to be used in multiple future vehicle models.
Figure 5. Bottom View of the PEV Prototype Body Showing the Battery Compartment The Joule battery needs to provide 70kW power at peak but due to its ~99% efficiency does not produce much heat. However it still requires some ventilation, presenting an airflow challenge, as shown in 6. The air is channelled from the vehicle front to between the two battery packs and flows outwards through the batteries to be vented at the rear. The battery itself is mounted on the closure panel (Figure 7) and is itself structurally stiff.
Figure 6. Air-flow Through the Battery Compartment
Figure 7. The Two Halves of a Joule Battery on the Battery Closure Panels In summary, a custom mechanical architecture is clearly essential for an EV.
Vehicle Electrical Architecture As vehicles have become more sophisticated, the percentage cost of the on-board electronics has grown exponentially. Taking out the engine and replacing it with a computer-controlled electric motor is a major inflection point, not merely an evolutionary change. As described earlier, the traditional approach to vehicle development is largely bottom-up, particularly for the electrical systems. The vehicle developer could select an “ABS system”, an engine (complete with Engine control Unit and electronics), electric windows (with a controller), an “alarm system” etc. These are standard offerings that are considered mature, which are then integrated with each other. This integration often results in a mixed bag of communication protocols, functional duplication and inefficient wiring design. If one considers that an EV is actually and electric device at heart, and then applies top-down functional analysis principles using available technologies, significant optimisation and opportunity arises. A central controller, with distributed nodes to all ancillaries, could for example allow the lights to flash when the charger is plugged in; a computer voice could provide audible vehicle warnings through the sound system; the interior heater could be turned on using your mobile phone; the vehicle could automatically be disabled if taken outside a designated geographic area; and all of this can be configurable via software. These potential benefits were recognised early in the Joule development process, giving rise to the philosophy that “Joule is a computer on wheels”, much to the chagrin of the mechanical engineers. The Joule developers were not the only ones to recognise this technology shift in cars, with Negele reporting in 2006 that “to meet these challenges the BMW Group basically rethought its development processes for electrics/electronics in order to introduce a more stringent systems engineering approach and set up a change program with clear focus on the E/E system as a whole. This signified an important turnaround from a “component-oriented” to a “system-oriented” development process.” (Negele at al, 2006). The electrical architecture therefore advocates a System Engineering approach in order to successfully integrate the various major components and requires flexibility to access off-the-shelf parts (with their diverse interfaces and protocols).
The Need for a Clean Sheet It can be seen from the above that much like the mobile phone was a totally disruptive product that did not simply replace the functionality of a landline telephone; an EV presents a totally new paradigm of vehicle and requires a fresh approach to its development. If the novel features could be provided in a desirable package, at an acceptable price ahead of the competitors, Joule had a good chance of being truly disruptive and becoming a success. The challenge would be to establish a development process that would elicit these benefits whilst still building on the vast expertise and supplier base of the existing automotive industry.
Vehicle Development Process We have seen already that the traditional automotive development approach is largely an evolutionary one. The development process established for the Joule development, on the other hand, was initially a top-down Systems Engineering approach. This was later merged with the automotive one to form a hybrid process, as described below.
Traditional Automotive Development What is not apparent from the automotive development process shown previously in Figure 3 is the underlying component-focussed process shown in Figure 8. This parts-centred approach is driven as a procurement and quality-management process, where even seemingly small component cost savings are pursued vigorously to gain significant benefits from volume purchase agreements. During the four phases the parts are evaluated, chosen, incorporated into the vehicle design, and their production established.
Figure 8: The Parts-focussed View of the Typical Automotive Development Process
During the A-sample phase, various component candidates are evaluated for fulfilling a particular function. This is typically where new part types, such as the latest braking systems from competing suppliers, are evaluated and the most suitable one or two selected according to the concept requirements of the next vehicle model. The selected parts may not be fully mature, but by the B-sample phase several of these parts are tested on mule5 vehicles to confirm their suitability and performance. In the C-sample phase the parts are integrated into what is intended to be the “final design-intent” vehicle, and must be fully mature. Many samples of this zero-excuse part are required to produce vehicles in low-volume (~100) that will be used for Design Validation (or Engineering Qualification) and thus exposed to the accelerated life testing and in-vehicle durability testing. The cost must be known accurately and the part manufacturing must have been established. In the final D-sample phase, the part is actually a production part made according to the final manufacturing process, but perhaps still in pilot volumes. D-sample parts are used to commission the vehicle assembly line and its related logistic processes. The vehicles built from these parts are subjected to testing focussed on establishing the repeatability and quality of the manufacturing and supply chain processes. Some engineering issues also emerge but changes to part designs at this time are considered highly undesirable.
Joule’s Initial Development Approach At the time the Joule development started, the company did not have much insight into the automotive process described above. A tailored Systems Engineering process was developed by the author, taking into account the risks and opportunities that were apparent at that time. As various cost-reducing route options were evaluated and the organisation grew, the top-level development process in Figure 9 was established.
Figure 9. Initial Top-level Development Process for the Joule6 5
A mule vehicle can be any vehicle to which the test part if fitted for evaluation under road or specific test conditions. 6 The diagram has been updated to reflect the names of the prototypes and the terms “C-samples” and “D-samples” (which were not known at the time) as points of reference for later discussion.
The major process inputs are listed on the left, leading into two main activities: a Technology Development process for the in-house development of systems that are unique to EVs (and thus present a higher technical risk); and a top-down Product Development Process that comprises of several waterfall-type phases of increasing depth. When the own-developed EV technologies became mature they would be transferred to the Product Development stream to be incorporated in the mainstream vehicle design. The following prototype phases were defined: PT1 had various technical purposes but also needed to attract further funding. It comprised of the technology platform that integrated the EV technology concepts (battery, drive system and controls) and allowed the team to learn about them. It also was the opportunity to develop the vehicle styling and character by making a full-scale model of the Joule, establishing a conceptual technical and market baseline. PT2 (Phase 0) followed, providing the first vehicle prototype that incorporated the unique vehicle style into a functional body, interior and chassis using in-house EV technologies, all integrated into one vehicle. The Vehicle Technical Specification (VTS) was quite extensive by this time and from this point forward was linked to the User Requirement, which became more mature as the marketing team introduced the vehicle concept to the market. A major purpose of this phase was to establish the Functional Baseline of the vehicle, whilst establishing a robust vehicle costing model to maintain alignment between cost and performance. Various technology options were evaluated off-line in this and the next phase, through integration into “mule” vehicles. PEV (Phase 1) was envisaged to be one in which the overall technical maturity would grow and the in-house technologies completed. The all-important “feel” of the vehicle would have to be evaluated in market clinics with potential customers and media, requiring four of these vehicles to be built. They would require “customer-ready” finishes and representative performance so that the User Requirement and high-level Vehicle Technical Specification could be validated. These prototypes are shown in Figure 2 and were the last Joules built. P50k (Phase 2) was the industrialisation phase for the Joule. This was planned as a major project over four years, in which the PEV low-volume concept would have to be engineered into a production-ready, market and cost-competitive product. One final design iteration was planned which would culminate in engineering qualification tests of around 70 vehicles in the field undergoing various tests. The production data-pack and supply chain would also be established in this time, as would the assembly plant and sales/support systems. The production qualification would be a reduced set of tests, prior to ramping up to full-scale production. Phases 3 and 4 would be production, supply, maintenance and ongoing improvements and “face-lifts” through the vehicle’s life-cycle. To execute Phase 2 would require the effort of a multi-national team of several hundred engineers and tens of component and system suppliers. As these relationships were established in Phase 1 and the detailed plans formulated, it was quickly apparent that the planned process and its Systems-Engineering terminology was a barrier to the required collaborations. A new process evolved, merging the automotive process with the original one.
Merging Two Process Views One of the shortcomings of the standard automotive process was the limited connection between the parts/system development and the vehicle development. Some of this activity should be top-down, yet there is an element of independent parts/system development required to address part-specific risks, complexity and life-cycle. The solution adopted by the company was to allow semi-independent processes for parts and the vehicle, with only a loose linkage, up to a specific convergence point. Figure 10 is modified version of Figure 9, showing the component development stream and how this converges with the vehicle development to form the Allocated Baseline of the vehicle. Modern engineering tools such as CAD, PLM, modelling and simulation could be employed for interface and requirements management prior to the convergence, with only limited physical integration and testing in a vehicle platform.
Figure 10. Convergence of New Technologies, Carry-over Parts and Vehicle Development This principle led the integrated Development Process shown in Figure 11. This funnel-like process provides a map of the relationships between vehicle development, system development, component development and assembly plant development. The top-down vehicle process is seen to start in the top-left while the parts process starts in the bottom left. The parts are selected through the A-sample and B-sample phases previously described, influenced by the early vehicle requirements. Unique EV-systems are also developed and evaluated in a similar fashion, but with greater strategic direction from the top-down vehicle process.
Confirm & Update VTS
‘A’ Virtual Vehicle
Confirm & Update
C Vehicles build
Engineering Qualification testing 1st Car Prodcution
‘B’ Virtual Vehicle
Confirm & Update ‘C’ Virtual Vehicle – Confirm & Update Component Detail
Confirm & Update Build and validate final design parts
Integ & Test Systems CTS
Confirm & Update
Lab & mule testing of parts
Systems & Parts
Prod Pre Prod Cars Ramp P1, P2 & P3 up Production Qualification testing
SOP + 3
Assy plant design/build
KEY to work streams:
URS – User Requirement Specification VTS – Vehicle Technical Specification STS – System Technical Specification CTS – Component Technical Specification PTS – Assy Plant Technical Specification
Figure 11. The Final Development Process for the Joule The parts and system requirements and interfaces are integrated in the “A-virtual vehicle” and “B-virtual vehicle” respectively to help confirm and update the User Requirement Specification (URS) and establish the Vehicle Technical Specification (VTS). Some of these virtual vehicles (like the PEV) may actually be built to help validate the requirements and integration concepts. The parts may undergo a lab or mule testing programmes and be integrated into systems, independent of the vehicle development stream, up to the Concept Evaluation Review (CER) convergence point, when the vehicle concept is frozen in the Allocated Baseline. After this the “C-sample” phase starts and the vehicle and parts processes must remain firmly in sync with each other and with the establishment of the procurement and assembly processes. Up to the CER there is an iterative process to link the entire organisational strategy, not only the vehicle design. The procurement of parts, their manufacture, the assembly plant, the market strategy, budget and project plan, all need to align and end up manufacturing, selling and supporting a vehicle which is a success in the market. This “Allocated Baseline” is thus very comprehensive, aligning all these aspect and becoming the major document governing the company activities going forward. Table 2 shows the progressive nature of this baseline development. “Book 1” would be completed by the start of the B-sample phase, whilst “Book 2” would be a more mature and prescriptive version, available at the CER.
Table 2: Gradual Establishment of the Vehicle Baseline by the Convergence Point Book 1 Book 2 Purchasing and Logistics A description of the strategy Detailed plan to handle the outlining the procurement & implementation of book 1 Strategy logistics concepts
Aftersales Strategy Quality Strategy
Plant concepts Plant technical Specs (PTS)
Vehicle concepts Vehicle technical Specs (VTS) Program Master Schedule (PMS) User Requirements (UR)
A description of the plan and policies relating to after-sales products and components. A description of the overall plan relating to the company-wide rollout of advanced product quality management A projection of the project cost breakdown and financial structure, from material to assembly, sales and support. A collection of individual design concepts that contribute to the production plant design An initial description of the specifications required for the whole assembly plant permanent features. A summary of the engineering concepts for the vehicle and parts. A set of technical vehicle design specs. Incl. mass & power budget, costed BOM, space allocation & technical standards. Compilation of individual sub-projects work breakdown structures for the project duration Collection of desired vehicle properties & features (“Voice of the Customer”).
concepts, based in the envisaged business framework A mature after-sales strategy reflecting changes made during the book 1 and 2 phases. A matured plan relating to APQP and rollout with regards to OE and its suppliers. An advanced model of the project cost, with accurate individual breakdowns per operation. A collection of matured plant concepts that contribute to the production plant design A mature list of specifications which can include modular specs which change with the vehicle model. Concept definition complete and feasible engineering concepts validated for the whole vehicle. Any changes made post book 1 phase to the UR will be reflected in this version of the VTS An updated version of the PMS reflecting all changes during book 1&2 activities. UR also reflects technical realities and cost compromises.
Conclusion The development of an EV from a clean sheet presents a significant challenge, particularly to an engineering team that has no automotive development experience. It does however also present a great opportunity - there are several unique EV benefits that can only be achieved through a fresh start of a vehicle’s design without the encumbrance of legacy designs. A fresh start also benefits more from a Systems Engineering approach, particularly when the legacy approach is to make only incremental improvements upon a prior product. The fresh approach provides an opportunity to apply Systems Engineering best practices in a tailored top-down process. Although many benefits were realised in this approach, it made it difficult to profit from the wealth of knowledge that is already available and the savings potential of incorporating off-the-shelf parts. A hybrid development process was therefore developed that merged the traditional automotive and Systems Engineering approaches.
The Joule project was halted during the “convergence point” for funding reasons, and never continued into the “C-sample” phase, preventing a test of the Joule Development Process. Up to that point the project was technically very successful, having achieved road-going prototypes within in a record time and low budget. The development process had passed the scrutiny of international partners and the Allocated Baseline was clearly established. It is hoped that the reader may have learned some lessons of their own in reading this paper, that may be applied to the benefit other complex product development projects.
References Oosthuizen, H. 2011. “Nothing Ventured, Nothing Gained.” Car Magazine, April 2011. Loureiro, G. Leaney, P.G. Hodgson, M. 1999. “A Systems Engineering Framework for Integrated Automotive Development.” INCOSE International Symposium, June 1999. doi 10.1002/j.2334-5837.1999.tb00264.x Magna Steyr. 2015. “Magna Steyr Product Development Process.” https://www.ecs.steyr.com/Product-Development-Process.1329.0.html?&L=1 Negele, H. Schmidt, R. Finkel, S. Wenzel, S. 2006. “Lessons Learned from Synchronizing Complex Systems Development within Automotive Industry.” INCOSE International Symposium, July 2006. doi 10.1002/j.2334-5837.2006.tb02770.x Swart, G.P. 2015. “Innovation lessons learned from the Joule EV Development.” Proceedings of IAMOT 2015 Conference, P235, http://www.iamot2015.com/2015proceedings/documents/P235.pdf Capgemeni Consulting. 2014. “Tesla Motors: A Silicon Valley Version of the Automotive Business Model.” http://www.slideshare.net/capgemini/tesla-motors-a-silicon-valley-version-of-the-automotive-bu siness-model?qid=86850eda-d7ac-4cfd-8cd7-13241dcc8cc9&v=default&b=&from_search=12
Biography Gerhard Swart is an experienced Systems Engineer that played a technical leadership role in various projects such as the Rooivalk Attack Helicopter, Hong Kong International Airport and the Southern African Large Telescope. In 2005 he was one of the founders and CTO of Optimal Energy. Today Gerhard is a Director of Alphadot (Pty) Ltd, which consults to government, industry and academia in Innovation, Product Development and Systems Engineering. He is also co-founder and CTO of BattCo Energy Storage Systems, who are commercialising battery systems for the African market. He is a registered Professional Engineer, member of INCOSE and senior member of SAIEE.
Acknowledgements The development of the Joule was an incredible journey by a team of passionate people. The dream was a lot bigger than one organisation and inspired many to hope against all odds, that they would make a significant impact on South Africa’s industry landscape whilst creating thousands of manufacturing jobs. Thank you to the team for bringing it so far. Thank you also to my saviour Jesus Christ, whose strength and encouragement helped me prosper through the journey.