The Invisible Conversation:
How Your Car Talks to You
Modern vehicles are engineered to be quieter, smoother and more automated than ever. But 15 years of peer-reviewed research reveals a troubling truth: in silencing the car, we may be silencing a safety-critical dialogue between driver and road.
Based on Vehicle Feedback and Driver Situation Awareness by Walker, Stanton & Salmon (CRC Press, 2018) — 15 years of funded research involving Jaguar, Ford, the UK EPSRC and the Institute of Advanced Motorists.
Do you remember someone — a parent, a driving instructor, an older colleague — telling you about their first car? How 70 mph really felt like 70? How the steering fought back in corners? How the engine screamed on the motorway?
Most of us dismiss these stories as nostalgia. The past was loud and uncomfortable; the present is refined and quiet. Progress, surely.
But what if that discomfort was the point? What if, hidden inside all that noise, vibration and mechanical resistance, there was a continuous, nuanced, safety-critical conversation between the car and the driver — a conversation that modern engineering has almost completely silenced?
That is the central, disturbing, and fascinating thesis of Vehicle Feedback and Driver Situation Awareness by Walker, Stanton and Salmon. Over 15 years of funded research, the authors built an airtight scientific case for something every experienced driving instructor has always sensed but never had the data to prove.
The car is not just a machine. It is a communication system. And we have been systematically destroying its ability to communicate.
The Performance Paradox
The numbers are staggering. In 1979 the Porsche 911 Turbo — the fastest production car ever tested at the time — produced 300 bhp and hit 60 mph in 5.3 seconds. Road testers described it as “startling” and unlike anything they had ever driven.
In 2018, a Volkswagen Golf — a family hatchback bought in supermarket car parks — produced 306 bhp from an engine 40% smaller, hit 60 mph in 4.9 seconds, and returned over 40 mpg.
The Golf is faster than the Porsche. It uses half the fuel. It weighs more, handles better and is exponentially more reliable. And yet every experienced driver senses something is missing.
The original Austin Mini gave drivers headaches above 50 mph. The engine screamed at 4,200 rpm on the motorway. Road testers called it “slow, painful, awful for long distance work.” Its 2018 successor runs the same motorway at a relaxed 2,300 rpm, barely audible. The sensory information has been replaced with comfortable silence.
More performance and less sensation is not simply a comfort improvement — it is a fundamental change in the information environment available to the driver. The question is: does that matter for safety?
The Airbus Warning Nobody Applied to Cars
The authors describe a trend they call the Eurofighter Effect — borrowed from aviation. The Eurofighter Typhoon is intentionally aerodynamically unstable. Without computer assistance it cannot be flown. Modern cars have not gone quite that far, but they are heading in that direction: drive-by-wire throttles, variable-ratio electric steering, and active suspension all interpose layers of computer authority between the driver and the road.
But the more disturbing comparison is with Airbus. When Airbus introduced fly-by-wire, they made a deliberate design choice. Unlike Boeing — where both pilots’ control columns are mechanically linked, so that if one pushes, the other feels it — Airbus used independent sidesticks with no force feedback and no mechanical connection between them.
The aircraft does not need this feedback. Technically there is no requirement for it. But humans do.
Air France 447 crashed in part because the captain pushed his sidestick forward while the co-pilot pulled back, and the aircraft averaged both inputs. Neither pilot knew what the other was doing — because there was no feedback path.
Salmon et al., 2016A NASA study of 164 Airbus pilots found 133 instances of “automation surprises” — moments when the automation did something the pilot did not expect — most arising from a lack of feedback.
Drivers are exquisitely sensitive to vehicle feedback — changes in vibration as small as 6% are detectable. Yet modern vehicle design increasingly removes the very signals that drivers’ nervous systems are tuned to receive. The panoply of advanced systems will exert an influence far greater than the sensitivity threshold of drivers to detect it.
The Science of “Feel”
The book defines vehicle feedback with scientific precision across five major systems. Each is a channel of communication between the road and the driver’s mind.
1. Powertrain Feedback
In an old carburetted engine, pressing the accelerator was a direct, proportional, immediate relationship. You learned by feel exactly how much pedal meant how much acceleration. In a modern turbocharged engine with drive-by-wire throttle management, the ECU mediates this relationship entirely. The feedback loop that taught drivers throttle sensitivity has been erased and replaced with a smoothed, comfortable approximation.
2. Transmission Feedback
The manual gearbox is a masterclass in haptic communication. The gear lever “affords” pushing and pulling. The clutch provides progressive resistance that tells you exactly where the engagement point is. If you try to engage a gear at too high a speed, the lever resists enormously. The gearbox talks. A modern dual-clutch automatic is faster and more efficient — and completely silent. The conversation is over.
3. Braking Feedback
The motoring press has generated approximately 400 distinct adjectives to describe brake feel: sharp, wooden, spongy, progressive, dead, bite-y. Each adjective represents a subtle deviation from linearity that a driver can detect and interpret as information about available grip and vehicle state. ABS systems interrupt this feedback loop deliberately — but the driver pressing the pedal to the floor while the ABS chatters is receiving profoundly confusing information.
4. Steering Feedback — The Most Critical Channel
Steering feel arises from aligning torque — the force the road and tyres exert back on the steering wheel. The harder you corner, the faster you travel, the more load the tyres are carrying, the more resistant the steering becomes. This resistance is a direct, real-time readout of tyre grip — the single most important safety variable in vehicle dynamics.
When steering suddenly goes light, it means the tyres have lost traction. When it builds progressively, grip is increasing. Experienced drivers read this signal continuously, often unconsciously. Electric power steering can reproduce a simulation of this feel — but many manufacturers tune it to feel consistently light and effortless regardless of load. The signal has been replaced with a comfortable lie.
5. Auditory and Tactile Feedback
This is where the research becomes most striking. Drivers who hear quieter engine noise consistently choose to drive faster than those hearing louder noise (Horswill & McKenna, 1999). Reduced auditory feedback leads to reduced headway and more risky gap acceptance. And perhaps most surprisingly:
Completely non-sighted individuals could estimate Time To Collision using auditory cues just as accurately as sighted individuals using visual cues.
Schiff & Oldak, 1990The ears are a genuine safety system, not a comfort feature. Every sound-deadening panel, every exhaust silencer, every active noise cancellation system removes signal that the driver’s brain was designed to use.
Situation Awareness — What the Driver Actually Needs
The book is built around the concept of Situation Awareness (SA), originally developed for military aviation. SA operates at three levels:
- Level 1 — Perception What is present in the environment right now? Speed, distance to hazards, road surface condition, vehicle state. Raw data. This is where vehicle feedback lives.
- Level 2 — Comprehension What does it mean? The car ahead is braking; this corner is tightening; I am going too fast for the conditions. Pattern recognition applied to the raw data.
- Level 3 — Projection What will happen next? At this speed I cannot stop in time. The road narrows ahead. This is the gold standard of expert driving — the ability to act before events force you to react.
Vehicle feedback is Level 1. Every piece of auditory, tactile, haptic and steering information the car provides is raw data for the driver’s awareness model. Strip it away and the driver is building their picture of the situation on incomplete data — like a pilot flying on instruments with half the gauges blacked out.
Levels 2 and 3 are only achievable when Level 1 is rich, accurate and continuous. This is the foundational link that the research establishes with scientific rigour.
A full task analysis of driving identified over 100 discrete information requirements across all driving tasks. Many of these — speed estimation, surface condition, vehicle loading, following distance — are primarily supplied by non-visual feedback. Removing them does not remove the requirement; it removes the driver’s ability to meet it.
The Ironies of Automation
Lisanne Bainbridge’s landmark 1983 paper defined irony as “a combination of circumstances, the result of which is the direct opposite of what might be expected.” Her ironies of automation apply perfectly to modern cars.
The authors built a custom driving simulator — a real Ford Mondeo around a high-resolution projection system — with independently switchable feedback channels: visual, auditory, steering force feedback, and seat vibration. Thirty-five drivers completed identical road scenarios across eight different feedback configurations.
Using Signal Detection Theory — the same mathematical framework used to assess radar operators and medical diagnosticians — they measured each driver’s sensitivity to safety-relevant events.
From visual only (complete sensory isolation) through to visual + auditory + steering + seat vibration (full feedback). Each condition was measured against the driver’s actual accuracy in perceiving SA-relevant events.
What the Data Showed
- In the visual-only condition — the modern “comfort bubble” — drivers were measurably, statistically worse at detecting safety-relevant information.
- Adding auditory feedback alone: situation awareness improved significantly.
- Adding steering force feedback alone: situation awareness improved significantly.
- Adding seat vibration alone: situation awareness improved significantly.
- With all four modalities combined: situation awareness was highest of all conditions.
Drivers in the sensory-deprived condition did not know they were worse. Their self-reported awareness remained roughly constant regardless of how much feedback they received. They felt just as aware. They weren’t.
This is the most dangerous finding in the book. Drivers in modern, well-isolated, highly automated cars feel perfectly in control. They are not perceiving the degradation in their situational awareness — because the very systems that would tell them about it have been removed.
Cars vs. Motorcycles — A Natural Experiment
One of the most elegant sections of the research compares car drivers and motorcyclists performing identical journeys over the same road course. Motorcyclists have no metal shell, no sound insulation, no power-assisted controls, no isolating suspension. The feedback is raw, continuous and inescapable.
Motorcyclists are not naturally more intelligent or more attentive than car drivers. They are receiving more feedback. And that feedback is building a richer, more accurate, more integrated mental model of the driving situation.
Structural analysis of their thinking — using propositional network analysis and graph theory — showed that motorcyclists’ awareness of the traffic situation functions more like a well-connected knowledge graph and less like a linear list. When something unexpected happens, they can traverse their mental model quickly and reach accurate conclusions.
Modern car design is systematically pushing car drivers towards a situational awareness profile that is structurally inferior to that of motorcyclists — not because car drivers are less capable, but because their vehicles are engineered to provide less information.
Advanced Driving — The Proof
A study of advanced drivers — trained to police or Institute of Advanced Motorists (IAM) standard — compared against ordinary drivers answers a fundamental question: does training change what drivers know, or just what they do?
Advanced driving in the UK and Ireland is based on the System of Car Control — a framework requiring constant, systematic observation, information gathering and anticipatory action. Good advanced drivers are not faster; they are earlier. They identify hazards at greater distance, process more information per mile, and make fewer reactive corrections because they have projected further ahead.
What the Research Confirmed
- Advanced drivers produced more information objects than ordinary drivers — they perceived more of the same environment.
- Their semantic networks had more novel, unique information elements — they comprehended the road differently.
- They gave greater importance to elements that ordinary drivers perceived but dismissed — they understood the significance of signals others filtered as irrelevant.
- Their knowledge networks were denser and more interconnected — a richer internal representation of the same driving situation.
This last finding is perhaps the most profound. The advanced driver does not just perceive more. They have a cognitive schema — built through training — that tells them what to do with the information they receive. The feedback from the steering wheel is not just a sensation; it is a data point in a deliberate, structured analysis of the road situation.
Advanced driving training rebuilds and enriches the Level 2 and Level 3 SA processing that converts raw vehicle feedback into predictive knowledge. It is not about faster reactions. It is about never needing them.
What Good Vehicle Design Should Look Like
The conclusions of 15 years of research are distilled into actionable design principles. These are not abstract academic recommendations — they are practical engineering directives with direct implications for every driver, instructor and fleet manager.
- Feedback is a system, not a feature. Vehicle feedback must be considered holistically across all modalities: visual, auditory, steering, tactile. Optimising one while eliminating another creates dangerous gaps in driver SA.
- Less is not more. The comfortable assumption that removing stimulation reduces cognitive load is false. Removing informative stimulation reduces SA. The two are not the same thing.
- Don’t confuse preference with need. Drivers prefer quiet, comfortable, effortless vehicles. They also need rich, accurate, continuous feedback. These desires are in conflict. The appropriate response is informative design — not elimination.
- Advanced automation requires enhanced feedback, not reduced feedback. As vehicles become more automated, the moments when the driver needs to intervene become rarer but more critical. SA at those intervention moments must be higher than ever. Yet automation, by its nature, degrades SA over time.
- SA requirements must drive design. Before any vehicle control system is designed, a full analysis of what information drivers need — at what level of SA, across what road types — should define the feedback specification. Currently, this is almost never done.
What This Means for Driving Instructors
The research has direct, practical implications for anyone who teaches or coaches drivers. The “sensory bubble” created by modern cars is not hypothetical — it is the everyday experience of every learner driver sitting in a current-generation vehicle.
Key Points for Teaching
- Feedback deprivation is invisible. Students in modern cars feel confident but may have significantly degraded SA. Do not rely on their self-reported awareness.
- Teach students to actively listen to and feel the car. Encourage motorcycle-style attentiveness even in a car cockpit.
- Auditory feedback is a speed regulator. Explain to students why engine noise matters and why electric vehicles require heightened visual compensation.
- Steering weight changes encode grip information. Never allow students to ignore changes in steering feel — make it a deliberate observation point.
- Advanced training rewires the SA network. It changes what students do with feedback, not merely whether they perceive it. This is the goal of good instruction.
- Overcautious use of driver aids can be as dangerous as their absence is uncomfortable. Teach students when systems are active and what that means for their awareness.
The Conversation Worth Having
There is a question worth sitting with. Globally, approximately 1.35 million people die in road traffic accidents each year. The majority of serious crashes involve a failure of situational awareness — not recklessness, not mechanical failure, but a driver who did not know what was happening around them until it was too late.
For decades, vehicle engineering has pursued a trajectory of isolation: quieter, smoother, more automated, more disconnected from the raw data of the road. And for decades, driving instructors, advanced drivers, traffic police and safety researchers have sensed — without always being able to prove — that something important was being lost.
This research proves it.
The car has always spoken to us through its vibrations, its resistance, its sounds and its responses. That language, built over a century of mechanical refinement, encodes real-time information about tyre grip, speed, hazard proximity, road surface and vehicle loading that no screen, no alert and no automation can fully replace.
The most dangerous mile may not be on a winding mountain road. It may be on a familiar commute, in a quiet, modern, perfectly comfortable car, driven by someone whose situation awareness is being quietly, efficiently and invisibly starved of the information it needs.
When we silence the car, we silence the conversation. The car is talking to you. The question is whether we — as engineers, instructors and drivers — are still willing to listen.
Build Your Situation Awareness
Advanced driving training develops the exact cognitive skills this research identifies — richer perception, deeper comprehension, and the ability to project and plan ahead. Book a lesson or explore our free resources.
Book online