Already in 2010, the sound design of electric vehicles was a topic at the Audio Branding Congress. This is an article of Markus Bodden and Tobias Belschner that was published in the proceedings of the Congress, the Audio Branding Academy Yearbook 2010/2011.
1. Introduction
Electric vehicles open a new domain for automotive industries. This is especially true for sound – they are extremely quiet at low speeds and they sound different. Their quietness poses a big risk for pedestrians and cyclists – the accident rate for them is double as high as for conventional vehicles. Electric and hybrid vehicles thus require a specifically generated external sound. But what should they sound like?
So far, there is practically no experience with synthetic vehicle sounds. In other domains, such as handy ringtones it has become evident that poor sound design can quickly deteriorate into acoustic chaos. The approach implemented in the ELVIS3 E-motion system is not only to satisfy the security aspect, but to tightly integrate security into a complete brand sound approach. The electronic vehicle needs an acoustical identity which has not been defined yet. In addition, the customer needs an emotional link to his car and has to experience the appropriate feedback.
To implement this, we have developed the Interactive Sound Signatures, an acoustic fingerprint of the overall sound. These base patterns are customized by the actual and dynamic vehicle parameters like speed, load, recuperation, gear etc. to form an interactive live sound. Due to the flexible architecture the process of finding the corresponding sounds is easily accomplished
2. History and prognosis of EV development
Electric vehicles have a strikingly long history – they were even developed and brought to market before vehicles with combustion engines. In 1881, i.e. 130 years ago, the first electric driven vehicle was presented at the International Electricity fair by Gustave Trouvé. A year later, Werner Siemens presented an electric powered coach, followed by several other approaches and implementations for electro-mobility.
It is interesting to note from today’s technical discussions that in 1900 the Lohner-Porsche was presented. The outstanding feature of that car: it had wheel hub motors, a technique where the motors are placed directly at the wheels, i.e. where the power is actually needed. Today’s models do not yet use this “advanced” technique – they still have a central electric engine directly replacing the former combustion engine. Nevertheless, this technique is considered as the technique for the future.
In the period up to 1940, electric vehicles achieved their maximal dispersion, being at percentages of up to 50% of the vehicles, e.g. in New York in 1901. In the following years, electric motors were more and more replaced by combustion engines, and until today only very rare models survived on the market. The knowledge of the restricted oil resources and the implied need to develop vehicles driven by alternative energy led to a revival of the electric vehicles in automotive development.
So far, no mass production vehicle has found its way to the European market yet, and there is a significant delay in the market introduction of these vehicles. In 2009 just 162 electric vehicles were newly registered in Germany – a percentage of 0.04% of all newly registered vehicles – not even a niche group. In 2010 we are still waiting for the availability of mass production vehicles with some releases finally being announced for the end of the year. Nevertheless, the prognosis of electric vehicle market shares is still positive (see Fig. 1) The german ministry of transportation expects and plans to have 1 million e-vehicles on german roads until 2020.
Figure 1. Prognosis of electric vehicle market shares (in german). From: CAR University Duisburg-Essen
3. Sound of electric vehicles
Electric vehicles are effectively not much different from the vehicles we know today, with their combustion engine being replaced by an electric engine. But regarding the sound this means that the acoustical heart (and even the acoustical soul) has been taken out of the vehicle. The sound of vehicles has always been associated with the sound of combustion engines, and much less with other sound components, like the sound from tires and wind. The specific differences in the sound are:
Standstill /Idle
Idle: now absolute quietness.
Starting up: only startup of components is audible, unfamiliar
Driving exterior
tires: physically as before (relevant from medium speeds up)
wind: physically as before (relevant at higher speeds)
motor: significantly different
Driving interior
tires: physically as before, but more present (relevant from medium speeds)
wind: physically as before, but more present (relevant at higher speeds)
motor: significantly different
new components: new introduced sounds (e.g., converter etc.)
other components: physically unchanged but perception can be significantly different (lower overall sound level, spectrally different masking)
In summary this means that sounds and their perception are significantly changed compared to what we know and what we have learned for about 100 years during the development of vehicles with combustion engines. These changes of the sound have a severe impact on various levels which will be discussed in the next chapter.
4. Security
The impacts can be divided into several groups. The first and most essential group is security. Compared to the other aspects this is a hard factor. When electric vehicles are introduced and presented, one aspect is usually pointed out as being positive for all, driver and society. The electric motor itself is much more quiet than a combustion engine, which means that the vehicle sound level is lower for those speeds where normally the combustion engine dominates.
This is true for the low speed range for speeds up to around 50 km/h. Some measurements indicate that the difference in sound level can only be observed up to a speed of 20 km/h (see Figure 2), so really at very low speeds. But this is only true for constant speed drive, and not for the more relevant driving situations in cities, which are accelerations etc.The pervasive deployment of electric vehicles in cities will lead to a reduction of the overall traffic noise level and thus to a partial relief of city noise pollution. Furthermore, the interior sound level is also reduced, advertised as being more comfortable for the driver.
Figure 2. Passby levels of Hybrid and combustion engine vehicles as a function of speed (constant speed, no acceleration). From: Study Committee on Low Noise Measures for Hybrid Vehicles, etc., 29 January 2010
These aspects seem to indicate that the “silence” which is associated with electric vehicles is a positive aspect which should be maintained – although this is not really true, since the vehicles are not completely silent, but just more silent than conventional vehicles.
4.1 Is silence blessing?
Is that really true? What is the impact of that reduced exterior sound in daily traffic, for example? The problem arising is indeed that the vehicle is too silent – a positive aspect turns into having negative consequences. The sound does no longer fulfill its function! The function is to tell the pedestrians that there is a vehicle with a certain speed approaching from a certain direction.
We are trained since primeval times to use our auditory system as a warning and surviving system. In former times it was the enemy, the wild animal that we heard approaching from behind. In the industrial age it is the vehicle approaching from behind. And there is no means for us to adapt to this by using other senses – our vision is only active in a limited part of our surrounding, while the auditory system works all the time even with multiple sources and all around us.
A basic danger is that electric vehicles emit no sound at all in standstill. Since there is no combustion engine which has to run with a minimal rpm (idle), there is just NO sound at all. This means that an electric vehicle can start from nothing, without any hint given to pedestrians that this is about to take place. Furthermore, the acceleration is quite different as well. With combustion engines we first have a strong increase in sound, since we first press the pedal and then start to move. Electric vehicles start linearly, without any sound boost, with high power. Even if we did hear that it would be too late!
The national society of the blind in the US initiated a study on the impact of the missing sound feedback. The results of the study have been published by the US National Highway Administration (NHTSA) in September 2009. The study is based on a statistics of the accidents of pedestrians and cyclists with vehicles in specific city driving conditions, using the group of Hybrid vehicles (HEV) compared to vehicles with combustion engine (ICE). The main results are :
“A variety of crash factors were examined to determine the relative incidence rates of HEVs versus ICE vehicles in a range of crash scenarios. For one group of scenarios, those in which a vehicle is slowing or stopping, backing up, or entering or leaving a parking space, a statistically significant effect was found due to engine type. The HEV was two times more likely to be involved in a pedestrian crash in these situations than was an ICE vehicle.”
“In conclusion, this study found that HEVs have a higher incidence rate of pedestrian and bicyclist crashes than do ICE vehicles in certain vehicle maneuvers. These results should serve as a guide when designing future HEV pedestrian and bicyclist crash prevention programs.”
It is important to note that this study is purely based on hybrid vehicles, so that there is no guarantee that they were actually driven in electric mode. If the battery is too low these vehicles use their combustion engine even for the driving conditions investigated. For electric vehicles, the expected outcome may very well be even worse. In consequence severe and provocative questions arise which are related to the introduction of electric vehicles to our city traffic:
- What will happen with the image of e-vehicles if the accident rates rise and, e.g., children are hurt or killed?
- How will assurance companies react – will they raise the rates for e-vehicles?
- How will the driver react to the stress that is invoked by the need to take care all the time for pedestrians or cyclists crossing? He should experience a relaxation of stress due to the reduced interior sound level, but on the contrary his stress level will rise because he has to be extremely careful all the time and be aware of the possible consequences.
Figure 3. Typical risk scenario for pedestrians. Which car will start to move without any warning in the next second?
4.2 Legislation
Since the study of the NHTSA has proven a significant increase in accidents in the US, legislation has started to define laws requiring a minimal exterior sound level as a pedestrian warning. Japan will be the first country where such a law will be passed. In the US and Europe such laws are also in the process of being prepared.
The level curve represented in Figure 1 is used as a reference for the Japanese law proposal. The conclusion is that electric vehicles require an external sound device for speeds of up to 20 km/h. The noise level generated by the devices should not exceed the noise levels of vehicles with combustion engine. We have to keep in mind here that the level curve in Fig. 1 is valid for constant speed situations only, and that the difference in level is different for acceleration. There are no limits of levels specified for this condition in the actual available proposal.
While the level of the external sound can be clearly defined, measured and controlled, the type of sound is hard to specify. The recommendations from Japan include that typical warning signals (like sirens, chimes or melodies) should be deemed, as well as nature or animal sounds. The sounds should be conceived as being generated by a vehicle, and the sound parameters should represent the vehicle action. The sound shall be automatically altered in volume or tone depending on the vehicle speed for easier recognition of the physical status of the vehicle.
4.3 Solution proposals for security
Several solution candidates have been presented to solve the security problem of reduced exterior sound.
Reproduction of combustion noise
The first approach is the exact reproduction of the major sound component that is missing for electric vehicles – the sound of the combustion engine. This sound shows the advantage that there would be no need to newly adapt to it since it is well-known and everybody is used to it. But exactly this is the major drawback. There is a significant change in technology, and the “dirty” combustion engine is replaced by a “clean” electric motor.
The sound on the other hand still claims to be “dirty”. We would thus not only hide the technical evolution of that new vehicle generation, we would even try to reverse it. In addition, the exact reproduction of combustion noise is a technical challenge. The result usually is an artificially sounding re-make – although the intention was that it should sound as before. The reproduction of combustion noise thus should not be considered as a viable solution.
Use of the horn
The easiest and cheapest choice is to use an external sound reproduction device which is already integrated in the vehicle: the horn. There would thus be no effort, no space needed and no price to be paid for additional loudspeakers to reproduce the exterior sound. The major drawback of the horns is that they are not really loudspeakers. They can only produce specific tonal signals, so basically what we know as horn sounds today. Disregarding this drawback, this approach was presented as the solution for the Chevrolet Volt, resulting in the vehicle producing an intermitting warning sound in city traffic.
Despite the fact that the recommendation for the Japanese law deems this type of noise, we just have to imagine what would happen to our city soundscape if this solution became reality – a cacophony of car horns honking all over the city day and night. While certainly a hilarious scene to imagine, it should be clear that a myriad of honking vehicles roaring through neighborhoods will not exactly improve our city’s soundscapes. We hope that legislation will ban it.
Other solutions
There are several other solutions presented (mainly on youtube.com, see references) where other, usually more modern sounds are used. Typical sounds are those which resemble the turbine sound of jet planes. From the available material it is hard to state how suitable those sounds are in the real use of the vehicles. In the examples present on youtube the sounds seem to be mixed in a studio to the videos and not recorded in reality by the driving vehicles.
Impact on society
The discussion above shows that the sound not only has an influence on the accident risk, it also has an impact on society. The next generation of vehicles will change the soundscape of our cities, and this has an impact on society in general. We thus can not leave it to the car manufacturers to define properties of the sounds, or even open this field completely for individualization of exterior sounds.
5. Other impacts of missing and changed sounds
The sound level is not only reduced in the exterior, but also in the interior. But not only the level is changed, also its spectral composition. The low frequencies produced by the combustion engine are missing, resulting in a different masking. This means that sounds of the different components in the vehicle are much better and more clearly audible, and this usually reduces the perceived sound quality.
This is of crucial importance especially for suppliers. They have experienced an increased demand on the sound quality of their components over the last years, and now they face another huge challenge. The problem is that these acoustic requirements would significantly increase the costs for the components but also the weight of the vehicles. Usually acoustical improvements require the installation of additional damping materials, significantly increasing the vehicle’s weight. This stands in contrast with the necessity to reduce weight and to use lightweight constructions (it has to be kept in mind that the batteries already increase the vehicle’s weight).
An even more severe drawback is the missing feedback and reduced interaction with the driver. Powering on the engine only can be perceived as a starting of different components, and the driver does not get any audible feedback if the engine is ready to start – there is no idle sound anymore since the electric engine is not running in standstill. This also holds true for acceleration, the feedback differs significantly from what we are used to.
Furthermore, emotion is not really created and supported by the available interior sounds. The emotional appeal to the driver is missing. If we imagine how actively emotion is used for the marketing of vehicles today, we can imagine how big the impact will be for future electric vehicles. Today much effort is spent to provide the different brands and vehicle types with specific sound characteristics, thus realizing a brand sound. This will all return to blank for electric vehicles – they basically sound all the same, there is no differentiation with regard to the sound character. The electric vehicles have no acoustic identity up to today. To sum up, the impact of the reduced and changed sound on the different levels is as follows:
- Pedestrians: accident risk
- Driver: missing feedback/interaction, reduced Sound Quality, lack of emotion
- Manufacturer: missing identity, Corporate Brand Approach
- Suppliers: further increased requirements concerning the Sound Quality of components
- Society: interaction between sounds, meaning for Soundscapes
This shows that there is a much deeper impact than just the security aspect. Nevertheless, Sound Design can be used to address all these impacts and to implement a solution combining these aspects.
6. Sound Design: Identity, Emotion and Interaction
The basic requirement to the sound of electric vehicles is to ensure the security aspect. Since that requires an external sound reproduction device, and since in the interior the available audio system can be used, the technical basics are available to implement a unified Sound Design approach to tackle the different impact problems which arise from the changed sound. This means that an entirely new sound identity has to be created – a completely new situation for vehicle manufacturers.
Up to today, Sound Design in automotive industries mainly was related to the design of interior feedback sounds (e.g., direction indicator, light warning, seat belt reminder etc) or the shaping of the base vehicle driving sound. The latter means that there is a set of existing sounds (engine, tires, wind), which is produced by their function, and that only some aspects of these sounds can be changed. This is completely different now. Sound Design more or less starts from scratch. So what is the Sound Identity of electric vehicles? How can this be defined? Who should define it?
6.1 Who defines the new sound?
The automotive industries have understood to include the customer demand as the decisive driver for its acoustical developments. They intensively use customer clinics and psychoacoustics to include the customer perception and opinion to define target sounds (e.g., Bodden 2006). In the case of electric vehicle sounds the situation is somehow different. The customers can not express at this time what the optimal sound would be. There simply is no experience with this new type of sounds to be developed, so that they just have to judge based on their imagination.
As a consequence, the answers people give about what would be an appropriate sound for an EV differ from no sound to a new sound, but without direct hints to what this sound should be. It also has to be considered that the EV sounds should not be appealing just with short-term perspective. That is indeed a new quality for Sound Design. With music and Sound Design for movies the target is a bit different – there the effect is in the focus, and the sounds are not active or present all the time.
Now interesting sounds have to be created, but they still have to have the same effect after an all-day drive and after using the vehicle for several years. The long-term effect thus has to be in the focus of the development and evaluation of the sounds. This means that the Sound Design for electric vehicles basically has to start from scratch. This is a complicated task, but on the other hand this offers unknown possibilities for car manufacturers. They now have the unique chance to manifest their specific brand sound.
6.2 What should Sound Design solve?
The security aspect just is the basic need which has to be fulfilled, while the contribution with regards to the other factors defines the quality of the EV Sound Design. With an appropriate Sound Design thus also the aspects emotion, identity and interaction have to be tackled. This requires an integrated and basic Sound Design approach. The result will be a unique and specific sound for each EV type and manufacturer.
The Sound Design process should develop features which provoke associations related to the new technology. Key phrases are green technology, environmentally friendly, innovative, electric etc. The emotion and identity factors have to be newly defined. In today’s vehicle the main driver for emotion is sportiness, mainly defined by the characteristics of the combustion engine. Thus one way would be to try to develop new sound characteristics for sportiness, or, more advanced, to define a new emotional binding to the vehicle, e.g., CO2-saving, or smoothness of driving. The question here again is if the situation for actual combustion engines should be reproduced or new relations established.
Figure 4. Sound Design for electric vehicles.
Figure 5. Associations to be used for the sound of electric vehicles.
6.3 The System ELVIS3 E-motion
The previous discussions have shown that the process required for proper EV Sound Design is tedious without a known outcome. What is required today thus is a flexible working tool which allows for the generation and easy adaptation of different sounds and sound generation schemes. A tool which implements this is the ELVIS3 E-motion system (ELectric Vehicle Integrated Sound Signature System). The philosophy of this approach is that the EV sound can be described in form of a unique base sound character, the Interactive Sound Signature. This reflects the global behavior of known vehicles – the sound changes with the status of the vehicle, but its base character should always be identifiable.
A base requirement is that the sound should reflect the dynamic features of the driving vehicle, it should reflect its motion and status. The dynamic sound is thus composed from the Interactive Sound Signatures which are adopted by the actual dynamic vehicle parameters. The most important parameter is the vehicle speed, but also other parameters like the engine load, the steering angle, recuperation mode, driving mode etc. should be considered.
The standard way to make these vehicle parameters available is to use a CAN-bus interface which is included in ELVIS3 E-motion. The system can thus easily be integrated into a vehicle, and the sounds can be tested during driving. Besides that the system can also be used as a laboratory standalone tool, where either prerecorded CAN-bus data can be auralized with different Interactive Sound Signatures to allow for a direct comparison of different base sounds, or driven with an external pedal and steering set and using a vehicle drive model.
In the laboratory case the influence of the other vehicle sounds on the overall sound has to be considered. These sounds thus also have to be modeled and adopted to the vehicle parameters. The kernel of the system is the flexible sound generation. Different sound synthesis and sound generation methods can be used to create nearly any sound character. The exterior playback of the system can either be realized by external loudspeakers fitted to the vehicle or by vibrational excitation of the vehicle structure, e.g. of the hood.
The more channels are used here, the more effects can be auralized, e.g. steering angle and thus moving direction with two loudspeakers in front, additional effects with rear loudspeakers, high Sound Quality at low frequencies with a subwoofer. For the interior the interior sound system is used for optimum results, with the signals fed in by an adapter, or, if unavailable, separate loudspeakers have to be used. For the laboratory situation normal loudspeakers are used.
Figure 6. Scheme of the ELVIS3 E-motion system.
The ELVIS3 E-motion system thus can be used as a sophisticated tool to identify, develop and evaluate EV sounds, as well as for implementation in the vehicle with dedicated hardware.
7. Summary
The introduction of a different engine type, the electric engine, changes the sound of these vehicles with severe impact on different levels. First and foremost, the reduced exterior noise leads to a doubled accident risk for pedestrians and cyclists in city traffic. This problem will be tackled by implementing laws for a minimal exterior noise level which are already under preparation.
In addition, the main interior Sound Quality changes – the sound level will be reduced, but the different components in the vehicle will become audible more clearly and will create numerous sound issues. Even more, the heart and soul of the vehicle sound will be washed out, the sound looses its emotion and feedback and interaction quality. Sound Design for electric vehicles can solve all of these problems – but it has to be performed by properly considering the long-term aspects and the social impact.
For the first time in automotive development there is a chance to shape entirely new brand sounds as well as brand identities. This is a tough and interesting challenge for Sound Designers, especially in that context since now creativity, vision and profound professional knowledge have to be tightly combined. ELVIS3 E-motion provides the solid and flexible foundation for this new and exciting Sound Design phase.
References
Report of the National Highway Safety Administration (NHTSA), USA
Study DOT HS 811 204, September 2009. Incidence of Pedestrian and Bicyclist Crashes by Hybrid Electric Passenger Vehicles.
Bodden, M. (2006): Product Sound Design. In: sonambiente berlin 2006. klang kunst sound art, H. de la Motte-Haber et al (Hrsg.). Kehrer Verlag, Heidelberg.
Youtube.com: Fake engine noise makes electric cars safer.
http://www.youtube.com/watch?v=Sol2MpIy5Rg
Youtube.com: Nissan Leaf VSP Vehicle Sound for Pedestrians.
http://www.youtube.com/watch?v=DwPwx-YxIZM
Youtube.com: Toyota Prius Noisemaker “Approaching Vehicle Audible System”.
http://www.youtube.com/watch?v=3Vy42zphNp4