Human Factors (also called Ergonomics) is a scientific discipline studying the relationship between humans and the systems they use. It draws on psychology, physiology, engineering, and biology. In driving, it asks one core question: How do the limits and capabilities of the human body and mind interact with the vehicle and road environment — and how do we design for safety?

• Origins: Human Factors emerged during World War II when the US military found that highly trained pilots were crashing not because they lacked skill, but because aircraft cockpit designs exceeded human perceptual and cognitive limits. MIT has taught formal HF engineering since the 1950s.
• The three pillars of HF in driving: Perception — what information the driver receives from the environment (vision, hearing, proprioception). Cognition — how the brain processes that information, makes decisions, and manages workload. Action — how commands are translated into physical control of the vehicle.
• Why it matters for road safety: NHTSA (US National Highway Traffic Safety Administration) analysis of 5,470 crash investigations found that 94% of serious crashes involved human error as a critical reason. Vehicle failure accounted for 2%. Environment accounted for 2%. Improving human performance is the single biggest lever for reducing crashes.
• The HF approach does NOT blame drivers: It recognises that human performance has predictable, scientifically understood limits. When a system (road, vehicle, or task) demands performance beyond those limits, failure is inevitable — regardless of how careful or experienced the driver is.

Driving is estimated to rely on vision for 90% of the information needed to operate a vehicle safely. Understanding exactly how the eye works — and where it fails — is fundamental to understanding driving safety.

• Contrast sensitivity: The ability to distinguish an object from its background. A pedestrian in dark clothing on an unlit road at night may have contrast sensitivity so low the eye literally cannot distinguish them from the background — even with functioning headlights. This is a physical, not attentional, failure.
• Dark adaptation: Transitioning from daylight to darkness takes 20–30 minutes for full rod activation. During this adaptation period, night vision is significantly impaired. One flash of oncoming full-beam headlights at night resets this entire process — you may need 3–7 seconds to regain full adaptation (during which you drive essentially blind).
• The night driving paradox: Rods (night-vision cells) have no receptors in the central fovea — there is a rod-free zone at the very centre of your vision. To see a dim object at night, you must look slightly to the side of it (about 5–10°). Looking directly at a dim obstacle at night makes it disappear.
• Age and vision: By age 60, the pupil admits only one-third of the light it admitted at age 20. The lens yellows and loses flexibility. Glare recovery takes longer. Contrast sensitivity falls. These changes are gradual and often unnoticed — which is why regular eye tests are safety-critical for older drivers.
• Speed and visual field: At 40km/h the useful visual field is about 100°. At 100km/h it narrows to around 40°. At 130km/h, approximately 30°. The faster you go, the more tunnel-like your vision becomes — even with perfect eyesight.

Attention is not a single thing — it is a collection of different cognitive processes. Understanding how they work explains many seemingly inexplicable crashes ("I just didn't see it") and forms the basis for evidence-based safe driving practice.

• Selective attention: The brain actively chooses what to process. It cannot process everything simultaneously. Driving requires constant prioritisation — junction ahead, speed, mirrors, road surface, other vehicles. When a secondary task (phone, conversation) enters the queue, the brain de-prioritises some driving inputs.
• Divided attention: The ability to do two cognitive tasks at once. Humans CAN divide attention between tasks, but with a cost — both tasks are performed worse than either alone. The degree of cost depends on whether the tasks compete for the same cognitive resource (e.g., two visual tasks compete more than a visual task + a simple manual task).
• Sustained attention (vigilance): The ability to remain alert to infrequent, unpredictable signals over long periods. This is exactly what motorway driving demands — and it is the type of attention that deteriorates fastest. After 20–30 minutes of monotonous driving, vigilance drops sharply and hazard detection slows significantly.
• Inattentional blindness: A well-documented phenomenon where a driver looks directly at an object but does not consciously perceive it because attention is directed elsewhere. The visual input reaches the eye, but the brain's processing is occupied and does not register the object. The driver genuinely does not "see" it — it is not an excuse, it is a measurable neurological event.

Christopher Wickens' Multiple Resource Theory — a cornerstone of MIT's HF engineering curriculum — explains why some task combinations are more dangerous than others. It is one of the most practically useful models for driver training.

Mica Endsley's model of Situation Awareness (SA) — a central concept in MIT's Human Factors curriculum — describes the complete cycle of how a driver builds and maintains an accurate mental picture of what is happening around them. Loss of SA is a primary cause of serious crashes.

• Working memory (short-term): The scratch pad of the mind. Holds approximately 7 items (±2) for about 20–30 seconds. While navigating, driving, watching traffic, AND processing a conversation, working memory is heavily loaded. When it fills, earlier items are dropped — often the current road position or the upcoming junction.
• Long-term memory and schemas: Experienced drivers store "scripts" for familiar situations — approaching a roundabout, joining a motorway, following a lorry in the rain. These schemas run automatically, freeing working memory for novel hazards. This is the actual mechanism by which experience improves driving — not faster reflexes, but smarter memory use.
• Mental models and their failure: A mental model is your internal representation of how something works. Drivers develop mental models of their vehicle (how it handles), familiar roads (where the tight corners are), and other drivers (what their behaviour means). When reality deviates from the mental model — new road layout, unfamiliar vehicle, unexpected driver behaviour — reaction time increases dramatically.
• Familiarity and complacency: Driving a familiar route activates a stored schema and reduces conscious attention. MIT research confirms that drivers in familiar environments show measurably lower scanning frequency, higher phone use, and more frequent workload lapses — on roads they believe they know well. Familiarity creates false safety.
• Change blindness on familiar routes: Temporary road works, a new junction layout, a changed speed limit — these are frequently missed by regular users who run their established mental schema rather than actively perceiving the current environment. "But there was never a stop sign there before."

Fatigue is one of the three pillars of road crash causation alongside speed and alcohol. Unlike the others, it is invisible to observers and nearly invisible to the driver themselves — because fatigue impairs the very faculties needed to recognise it.

• Two biological drivers of fatigue: Sleep pressure (homeostatic drive) — adenosine, a waste product of neural activity, accumulates in the brain throughout wakefulness and creates increasing pressure to sleep. Sleep clears adenosine. After 16 hours awake, adenosine levels create significant impairment. After 24 hours, impairment is severe. Circadian rhythm — the brain's 24-hour clock independently drives alertness cycles, with minima at 2–6am and 1–3pm regardless of sleep history.
• Microsleeps: Episodes of 2–30 seconds of sleep-stage brain activity occurring while the person appears awake and may even have eyes open. The driver is completely unconscious during a microsleep — there is no partial awareness. At 100km/h, a 5-second microsleep = 139 metres of completely uncontrolled vehicle travel.
• The fatigue paradox: Fatigue reduces the ability to accurately assess one's own fatigue. Highly fatigued individuals consistently rate themselves as "a little tired" or "fine." This is why "I know when I'm too tired to drive" is one of the most dangerous things a driver can believe.
• Driving impairment equivalent: 17 hours awake = BAC 0.05% (legal limit in most EU countries). 21 hours awake = BAC 0.08% (UK legal limit). 24 hours awake = BAC 0.10% (above limit everywhere). This equivalence is based on psychomotor vigilance task performance, not a metaphor — the actual driving impairment is equivalent.

The suprachiasmatic nucleus (SCN) in the hypothalamus acts as the brain's master clock, controlling alertness, core body temperature, hormone release, and dozens of physiological cycles on an approximately 24.2-hour cycle, synchronised by daylight.

Nobel laureate Daniel Kahneman's Dual Process Theory — widely taught in MIT's cognitive science curriculum — provides the most useful framework for understanding driving decisions. It explains both expert driving skill and common crash causes.

• The experienced driver advantage: Expert drivers have built large libraries of System 1 patterns for complex driving situations. What a novice must laboriously work through with System 2 (approaching a complicated roundabout, judging a gap), the expert handles automatically — freeing System 2 resources for genuine hazards.
• When System 1 kills: In an emergency, System 1 responds first — often the wrong response (e.g., braking hard mid-corner instead of maintaining speed and steering). Emergency manoeuvre training works precisely because it replaces the wrong System 1 habit with a trained correct one through repetition until it becomes automatic.

The term "reaction time" is commonly used but poorly understood. What drivers and courts call "reaction time" is actually a chain of several distinct physiological and cognitive processes, each with its own duration and failure modes.

• The amygdala hijack: Under sudden, severe threat (child running into road, unexpected skid), the amygdala — the brain's threat detector — bypasses the prefrontal cortex (rational decision-making) and triggers an immediate fight-or-flight response. This can be faster than conscious thought but may be wrong for the driving context (freezing, or stamping the wrong pedal).
• Chronic stress: Prolonged stress (work pressure, financial worry, relationship problems) chronically elevates cortisol. High cortisol degrades working memory capacity, reduces patience, increases impulsivity, and narrows attentional focus. MIT AgeLab research directly correlated driver stress levels with road type and traffic density — city driving produces measurably higher cortisol responses than rural driving.
• Anger and aggressive driving: Anger activates the same amygdala pathways as acute threat. An angry driver shows: narrowed visual attention (misses peripheral hazards), faster and less accurate decision-making, reduced risk perception, increased tendency to interpret ambiguous situations as threatening ("that driver cut me off deliberately"). Anger is a genuine driving impairment.
• Running late: Time pressure increases risk-taking in a very specific way — it causes drivers to accept smaller gaps at junctions, overtake in shorter available distance, and reduce following distances. Studies show that the probability of a gap-acceptance error (pulling out in front of a vehicle) increases by approximately 15% for each minute a driver is "behind schedule".
• Emotional regulation as a driving skill: MIT's approach identifies emotional self-regulation — recognising and managing one's emotional state before driving — as a genuine safety-critical driving competence, not just a personality trait. It can be taught and practised.

MIT's manual control engineering lectures model the driver as a control system — receiving inputs, processing them, and generating outputs. This engineering perspective reveals exactly why smooth driving is safer and why control breaks down under certain conditions.

• The control loop: Driver perceives current state (position, speed, heading) → compares to desired state (lane centre, target speed) → computes error → generates correction (steering/throttle/brake input) → vehicle responds → driver perceives new state → loop repeats. This loop runs approximately 2–4 times per second.
• Preview and prediction: Expert drivers use a preview strategy — they look ahead and use the information to predict what corrections will be needed before the error occurs. This is feedforward control. Novice drivers use feedback control — they wait for an error to occur, then correct it. Feedforward is smoother, safer, and requires less corrective input.
• Control gain and instability: If a driver over-corrects (applies too much steering for a small error), the vehicle overshoots, a larger error occurs, the driver over-corrects again — and oscillation develops. On icy roads or at high speed, this oscillation can rapidly amplify to a loss of control. This is why smooth, proportionate inputs are not just comfortable — they are the correct engineering approach to vehicle control.
• Workload spikes: Most real crashes occur not because the driver is continuously inattentive but because their workload spiked at a critical moment — a junction, a pedestrian, a sudden hazard — while cognitive resources were depleted by a secondary task. The control loop missed one cycle and recovery was insufficient.
• Haptic feedback: Experienced drivers use tactile information from the steering wheel — vibration, weight changes, road feel — as additional input channels. This feedback is eliminated by distraction, gloves, or steering column rigidity. Disconnection from this channel reduces the driver's effective bandwidth.

The vestibular system (inner ear) and the somatosensory system (body position sensors) provide the driver with motion and orientation information. In most conditions this is reliable and helpful. In certain driving conditions, these systems can be actively misleading.

• The vestibular apparatus: Three semicircular canals detect rotational acceleration. Two otolith organs (utricle and saccule) detect linear acceleration and gravity. These feed continuous orientation data to the brain. In normal driving conditions on visible roads, they supplement visual information. In degraded visibility (fog, snow, night), they may be the primary source — and may be wrong.
• The lean illusion on curved roads: After sustained gentle cornering for 30+ seconds (e.g., a long gentle bend), the semicircular canal fluid reaches equilibrium and ceases signalling rotation. When the road straightens, the canal signals what feels like a turn in the opposite direction — the driver instinctively steers back into the bend. This is a documented cause of single-vehicle run-off-road crashes on long curves.
• Speed perception on motorways: After extended motorway driving at 110km/h, the brain adapts. Slowing to a slip road speed of 50km/h genuinely feels slower than 50km/h — the brain's reference point has shifted. This is why motorway exit crashes often involve drivers genuinely unaware they are still travelling at 80–90km/h on a 50km/h slip road.
• Cambered road disorientation: A road with strong adverse camber (tilting the vehicle away from the bend) can make the driver feel they are upright on a flat road. They may understeer because their vestibular system is not signalling what vision tells them — creating confusion about the appropriate steering input.
• The rule: When bodily sensation conflicts with visual and instrument information — always trust the instruments and visual references over the body's feeling. This is the same rule used in aviation since the 1920s.

• How alcohol impairs driving — precisely: Ethanol is a CNS depressant. At BAC 0.02–0.05% (below the legal limit in most countries): visual tracking impairment, divided attention deficits, slowed emergency braking. At 0.05% (legal limit Ireland and most EU countries): all of the above plus lane control degradation, situation awareness reduction. At 0.08% (UK limit): significant impairment across all driving tasks — hazard perception, speed estimation, gap acceptance, lane keeping. At 0.10%+: severe multi-system impairment.
• The risk curve is not linear: The relative crash risk at BAC 0.05% is approximately 1.5× sober. At 0.08%: 3–5×. At 0.10%: 5–10×. At 0.15%: 12–20×. At 0.20%+: 40×. The relationship is exponential — small increases in BAC above the legal limit produce dramatically increased crash risk.
• Cannabis: THC impairs tracking ability, divides attention, slows reaction time, and distorts time perception. Crucially, THC accumulates in fat tissue and impairs driving for up to 4 hours after a single use — and acute impairment may not correlate with perceived impairment. Cannabis users consistently underestimate their driving impairment.
• Prescribed medications: Benzodiazepines (sedatives, sleep aids): directly impair all driving-relevant cognitive functions. First-generation antihistamines: sedation and attention impairment equivalent to BAC 0.05% in some studies. Some antidepressants, painkillers, and muscle relaxants carry specific driving impairment warnings. The legal presence of a prescription does not make driving safe.

Hazard perception is the ability to read the road environment and identify situations that have the potential to require evasive action. It is one of the most reliably measurable driving skills and one of the strongest predictors of crash involvement.

• Types of hazards: Developing hazards — things that are already there but not yet threatening (parked delivery van, children at pavement edge, oncoming vehicle with high beam). Latent hazards — things that could become threatening depending on other events (ball near kerb, school gate, junction with limited visibility). Expert drivers respond to latent hazards; novices often only respond to developing ones.
• The 12-second scan: Expert drivers habitually fixate points 12–15 seconds of travel time ahead. At 60km/h that is approximately 200 metres ahead. This long-range scanning allows maximum time for SA Level 2 and Level 3 processing before a hazard becomes critical. Research at TRL (UK) and MIT confirms this as the single most differentiating eye-movement behaviour between crash-involved and crash-free drivers.
• Hazard perception is trainable: Multiple RCT (randomised controlled trial) studies confirm that hazard perception training — watching videos, identifying hazards, receiving feedback — significantly reduces crash rates in newly-qualified drivers over 2 years. The effect size is comparable to one year's additional driving experience.
• What expert scanners actually do: University of Southampton eye-tracking studies (replicated in MIT research) show that experienced drivers: fixate fewer but more relevant points; spend longer on hazard-relevant areas; begin braking earlier with smaller deceleration; and show wider spread scan patterns in urban areas and more focused central scanning on open roads.

• The design of displays: MIT HF principles for display design include: place critical information within central vision (within 15° of the driver's normal straight-ahead gaze); use analogue/colour coding for rapid pattern recognition (speed dials are faster to read than numerical displays); avoid cluttered information presentation that requires decoding; and ensure adequate contrast for all lighting conditions.
• Head-up displays (HUDs): MIT research evaluated HUDs extensively. They reduce eyes-off-road time for key information (speed, navigation turns) by projecting data onto the windscreen in the driver's normal viewing area. However, they can create cognitive capture — attention drawn to the HUD at the expense of the real world behind it.
• Touchscreens — the evidence: MIT AgeLab and AAA research measured interaction time with modern touchscreen infotainment. Tasks including entering a destination (40 seconds eyes-off-road equivalent), changing radio station (up to 20 seconds), and adjusting climate control (8–15 seconds) — all far exceeding the 2-second safe glance limit. Physical rotary knobs and buttons can be operated by touch without looking; screens cannot.
• Seating and posture: Correct seating position is an HF safety issue, not just comfort. Arms slightly bent at the steering wheel (not fully extended) allows fast full-lock steering. Seat back reclined less than 30° from vertical. Head restraint immediately behind the head (not below) reduces whiplash by 40% in rear impacts. Eyes approximately 7–8cm from the top of the windscreen opening.
• Mirror adjustment: Correctly adjusted mirrors eliminate the traditional blind spot at 7–9 o'clock (left) and 3–5 o'clock (right). The SAE blind-spot elimination method (mirrors angled out further than traditional rear-facing position) was developed from MIT/HF research into mirror field-of-view optimisation.

The practical translation of MIT's HF research into driver training produces specific, evidence-based interventions that are demonstrably more effective than traditional instruction-based approaches.

• Deliberate practice: Ericsson's research (incorporated in MIT's engineering curriculum) shows that expertise is built through deliberate practice — not just repetition, but targeted work on specific weaknesses with immediate feedback. A learner who completes 30 hours of driving without targeted feedback on their scanning pattern learns 30 hours of their own habits — not necessarily good ones.
• Metacognitive training: Teaching drivers to monitor their own performance — "Am I scanning far enough ahead? Is my gap sufficient? Am I following too closely?" — activates System 2 oversight of System 1 habits and is strongly predictive of long-term safe driving behaviour after qualification.

Every section of this presentation connects to a single truth: the most important safety system in any vehicle is the driver's brain and body. Understanding these systems scientifically transforms how we approach training, and how we approach every journey.