There are 3,000 casualties per year resulting from 2504 crashes in Melbourne in 2017. This included 104 fatalities, 269 cyclists and 397 pedestrians. Safety is measured reliably by serious injury crashes, requiring hospital admission. Crash data is from VicRoads.
People have the freedom to drive relatively dangerous cars, and motorbikes, with no automatic speed control, nor red light control, as is now proven possible with driver-less cars, to drive drunk, or under the influence of drugs, unlicensed or unregistered, with limited eyesight or medical conditions, to engage in high speed chases with police, and the existential rules are to give them a slap on the wrist, even for repeat offenders. But the 3,000 casualties is a significant risk that will never be acceptable, and would not be allowed in other disciplines, including the chemical industry and the legal system.
There are practical limits to the safety of driver-less cars. For perfect safety, driver-less cars can only travel at 10kph near pedestrians and less beside cyclists. This is not acceptable to car companies. Driver-less cars can not be complemented by driver-less cyclists and driver-less pedestrians, so exclusive track is necessary for safety. Driver-less cars require a driver for the foreseeable future. The safety benefits of driver-less cars are worthwhile, yet not sufficient. Any involvement of humans in road safety will always be a concern, and only fully automatic systems on exclusive track have proven to satisfy the most stringent safety standards, as exampled by Morgantown West Virginia group rapid transit and Ultra rapid transit at Heathrow.
Cars are an inalienable right of freedom and should not be taken away from people, so it is proposed to make alternative safer transport available and much more attractive. It follows that traffic must be attracted out of cars with superior service, based on cost, trip time, and reliability. Using a World Bank survey of price elasticities of demand, it is estimated that 80% is the maximum that can be achieved since some functions are best serviced by cars and trucks. Requirements: To achieve zero crashes, fully automatic vehicles on exclusive track are the only proven way. The benchmark is to attract 80% of traffic out of cars, on the bases of trip cost, time and reliability, and on to up-market fully automatic vehicles on exclusive track.
Requirements for system issues where car traffic must be reduced:
To attract 80% out of cars: speed must average 60kph, four times faster than cars, and six times faster than existing public transport; reliability must be not impacted by serial queues, traffic incidents nor maintenance; and low cost vehicles must weigh only 300kg, run smoothly, and cost 5c/vehicle-km, with two-way track cost of $13M/km, including stops and 100 vehicles. (This performance is available from possible options.) It means exclusive track, nimble, up-market, fully automatic vehicles, multiple sensors, control systems, interchanges, off-line stops, parallel parking, and off-line turns. All control and propulsion components and podcar networks must have redundancy. Route capacity must be one vehicle per second. Vehicles must be fail-safe for 300,000km. For the remaining 20% of existing traffic, safer 2-phase intersections, safer exclusive track for cyclists and pedestrians, and service roads as a buffer for abutting properties should improve road safety. Attracting 80% of traffic away from cars, onto safe transport, would save 2,000 casualty crashes per year, valued at ~$1B/year.