Time to hit the open source road
It's time for a radical rethink in the way we make, fuel and use cars. E&T takes a look at different approaches to designing eco-friendly vehicles using open-source principles and sustainable, proven technology.
It doesn't take a clairvoyant to see that the days of the petrol-fuelled car are numbered. Not only is the spectre of 'peak oil' - the point after which oil production enters terminal decline - looming, but some analysts are predicting a supply crunch when the world's economies come out of recession. Add to that the impact of car emissions on the environment and it's clearly time for a radical rethink in the way cars are manufactured, powered and utilised.
The major carmakers are only too well aware of this - witness Toyota's Prius and Honda's Insight as their response. But these are still the end-products of a mindset that dates back to Henry Ford, and which are still in thrall to a dwindling resource that has to be pumped out of the Earth's crust.
But now a number of groups around the world have begin to take a different approach, designing eco-friendly cars using open-source principles and sustainable, proven technology.
Anyone who uses the Mozilla Firefox Web browser or Linux operating system for example will have encountered the inherent advantages of open-source, whereby the software's original source code - the language in which the program is first written - is made freely available to users in the public domain, who can then modify and improve it, and in turn share those improvements with other users.
Firefox was recently reported to have passed the one-billion download mark, while by contrast, Microsoft's Internet Explorer - whose source code is not available to the public - has been dogged by security issues. And, while Linux has been growing in popularity in laptops, netbooks and even smartphones, Microsoft's Windows Vista operating system has been avoided by many companies and consumers because of problems with hardware and software compatibility.
What is an open-source car?
The drawback with traditionally marketed products is that users get only what they're given by the manufacturer, which may have its own commercial agenda. Microsoft, in this instance, is fiercely protective of its source code, releasing only the converted binary versions, which can be understood by computers but not people, preventing users from modifying them. But this isn't the case with open-source software and it won't be the case with open-source cars.
Granted, a piece of software is very different from a car, but by placing the cars' base designs in the public domain the developers point out that their approach and ethos is very much in the spirit of open-source software, even if they're not sure yet exactly how the development process will work.
And there's the rub: it's still very early days for these projects. At the moment there are three active efforts towards open-source cars - Riversimple in the UK, the Kernel Open Source Green Vehicle (OSGV) in the US and the c,mm,n (pronounced "common") project in the Netherlands. Yet, so far, only Riversimple has been able to show off a vehicle that's possible with current technology. And the widely accepted pioneer in this field, OScar, appears to have been moribund for some years. As a result, specific details about the control technologies in these projects are at times frustratingly vague.
Riversimple offers the clearest picture so far, however. It has designed a two-seat urban car - tentatively called the Hyrban - powered by a 6kW hydrogen fuel cell and 30kW of regenerative braking to charge ultracapacitors that feed electric motors on each of the four wheels, to deliver an estimated 300mpg (equivalent), a maximum speed of 50mph, a 200-mile range and emissions of 48g per mile. The designs, which are now with the 40 Fires Foundation - a forum to develop energy-efficient cars - describe a 'network electric' concept, where electric power can flow in multiple directions: from the fuel cell to the motors or ultracapacitors; from the motors to the ultracapacitors, and so on.
To optimise this energy flow, the control system, or vehicle system controller, has four primary functions:
- to interpret the driver's requirements and inputs, and so adapt to the style of the driver;
- to look after energy management, based on the state of charge and driving style, and allocate charge across the system;
- vehicle management coordination to allocate torque to the motors and manage the regenerative braking; and
- vehicle mode arbitration to handle the start-up and shutdown procedures.
CANbus off the shelf
The system is a blend of the standard and the bespoke. As Dr James Marco of Cranfield University's Automotive Studies Group, which developed the system, explains: "We wanted to prove the off-the-shelf technology, so we used CANbus around the vehicle linked to five local CAN slave controllers - the four motors and the ultracapacitor - as well as the fuel cell's own system and a standard automotive controller, a Pi Shurlock Open ECU.
"But we also designed the control system from scratch, exploiting existing research and extending other work at the university. As to whether the end results are open source, however, is up to Riversimple, although the control system has been designed to be modular, and in a sense it's already 'out there' because we've published academic papers on it."
Riversimple plans a production run of about 5,000-10,000 cars a year from 2013 for use in one or more UK towns, available on a lease-only basis and with infrastructure support from hydrogen producer BOC. More ambitious perhaps, however, are plans by US organisation the Society for Sustainable Mobility (SSM) for its Kernel OSGV.
The kernel Hybrid SUV
Americans love sports utility vehicles (SUVs) - a fact acknowledged by the SSM. Its plan is to make a seven-seat crossover SUV - the Kernel - again with an electric motor on each wheel, with a combined output of about 150kW, and, although it too will use ultracapacitors and regenerative braking, it will also feature lithium-ion batteries as standard in the base model as well as optional self-contained 'plug and play' power plants using petrol, biodiesel, hydrogen fuel cell or natural gas for example.
This base, or 'barebone', model is expected to do 0-60 in about nine seconds, reach a top speed of 120mph, return about 100mpg (equivalent) and have a range of about 30 miles on a single charge, which should be extendable to about 600 miles with an optional power plant.
But, while Riversimple is not talking up the four-motor aspect of its car, four-wheel drive is an important feature in the Kernel, which will also have drive-by-wire and a control system to adjust the power of each motor. And both point out that this set-up allows some elegant software solutions to the issues of stability and roll/yaw control - although, this being an open-source process, much of this is still in flux.
Production, with a run of about 10,000 units in the first year and rising to 50,000-100,000 units after that, is scheduled to begin in late 2010. And, while Riversimple's Hyrban is at the niche, 'smart car' end of the market, the $25,000-30,000 Kernel is aimed squarely at taking on the Toyota Highlander.
That puts the prospect of an open-source production vehicle actually on the roads very close at hand - a prospect boosted by the US's existing and growing electric charging infrastructure.
c,mm,n prototype at Autorai Amsterdam
Rather further off, however, is the c,mm,n project in the Netherlands, which is aiming to provide a model for cars by 2020. Although still at the concept stage, a four-seat prototype was unveiled at the AutoRAI car show in Amsterdam at the end of March.
Like the Hyrban, the c,mm,n car will use a hydrogen fuel cell battery and regenerative braking, and will also have a leasing option - but, similar to the Kernel, there will be the option of a range-extending diesel unit as well. Where c,mm,n differs is in its emphasis on autonomous driving and intelligent onboard systems that will allow the driver and passengers to read, check emails, even watch a film during the journey but still regain control at any time.
How this system will work in reality is unclear, as it's still at the simulation stage. But it won't be intended for driving in busy urban areas, as the project's simulation coordinator Sander Veenstra explains, "When driving in an urban area, the driver needs information on the speed and traffic lights to drive appropriately. On a highway this information is less important because the car decides on the speed.
"In our simulation, autonomous driving consisted of a lane-keeping system and a cruise control with a car-following system. As a driver enters the highway the car switches to the autonomous driving mode.
"The longitudinal behaviour of the car on a highway will be taken over by a cooperative adaptive cruise control, which will normally work as a cruise control with a set desired speed," he says. "But when an obstacle appears, the car will automatically brake to avoid a collision. Such a system is already used in the higher range Mercedes models. The cars will also be able to communicate with each other, for example by radio, about such things as speed or the traffic situation downstream, which will allow them to follow each other for short distances."
The car's range will be an estimated 95 miles but it's reported that most of the Dutch electricity networks have agreed to help roll out charging points across the country, which would extend the range indefinitely, and that they could charge cars in as little as 15 minutes.
Veenstra is keen to stress that the project is focusing on which features the car's users will want to see in it, and this is a common thread across each of these different projects - giving customers what they want but that also happens to be eco-friendly.
Also, while the c,mm,n project is concerned only with the Netherlands at the moment, implicit in the Hyrban and Kernel projects is the scope for their cars to be built as different variants around the world according to local user needs, and component and fuel supplies and infrastructures. This is timely, given that major manufacturers such as Boeing and Philips said recently that they are abandoning global supply chains in favour or regional ones in response to the credit crunch and concerns over climate change.
Some proponents of these open source projects say the big car makers have been too slow and done too little to address the problems of vehicle emissions and the wider sustainability issues of their products. Equally though, they accept the folly of trying to beat them at their own game. But now they've changed the game, and before long for many of us the cars we drive will be the ones we actually want, when we want.