- InsightsInsights
- CryptocurrenciesCryptocurrencies
- Stocks
- White Papers
- IndustryIndustry
- GeographyGeography
Last week, electric scooter giant Lime announced plans to trial a new bespoke computer vision platform to detect when users are engaging in the dangerous practice of riding on the sidewalk. Capable of alerting the rider to their transgression and potentially even slowing the scooter down, the safety mechanism was sorely needed, given that a spate of dangerous accidents has cast aspersions on the popular mode of urban transport.
It's not just among e-scooters that AI has a valuable role to play, either. Fatal railway accidents are a worryingly common occurrence, while road traffic-related incidents remain a leading cause of premature death worldwide, especially among younger people. Fortunately, AI- and computer vision-inspired solutions are surfacing which look set to enhance safety for a wide range of modes of transport, spelling excellent news for pedestrians, cyclists, scooter riders, motorcyclists, drivers, and passengers alike.
Lime Vision, claimed by its proprietors as the industry's first AI-enabled computer vision platform, is scheduled to be tested on almost 400 e-scooters in Chicago and San Francisco next month, with the trial rolled out to six cities by the end of the year. According to company president Joe Kraus, the camera-based technology underpinning Lime Vision is superior to rival GPS-focused platforms in its potential for other applications aimed at enhancing scooter safety.
Such innovations are highly welcome and perhaps even overdue for a mode of transport that was recently discovered to lead to more accidents than even motorcycles. According to a study conducted by UCLA, there are 115 injuries per million e-scooter riders. Among motorcyclists, the figure falls to 104 per million, while it's just 15 for cyclists. Scooters don't just pose a safety risk to their riders, either—their proliferation on sidewalks has created a daunting and potentially dangerous ordeal for the elderly, the visually impaired, and other vulnerable groups.
As one of the newest transportation options on the block, it's easy to pin blame on scooters – but transport, in general, could benefit from an AI-powered safety upgrade. The risks of rail travel were driven home late last month when two fatal crashes involving Amtrak trains occurred within days of one another. The first occurred in Northern California and claimed three lives, while the second took place in Missouri, killing four and seriously injuring around 150 more. Both incidents came at crossings without guard rails or lights, but the costs involved in implementing those kinds of safety features can be extremely prohibitive.
Road traffic is even more hazardous to human health. A recent UN report found that over 1.3 million people die from road traffic accidents a year, making it the leading cause of premature death among five- to 29-year-olds. According to a recent study, although road injuries and deaths have fallen by a modest degree over the last three decades in wealthy nations, a subsequent spike in rates in low- to middle-income countries (LMICs) – where 93% of deaths occur – has offset those improvements. As a result, the UN has pledged to halve those tallies by 2030.
Technology looks set to play a huge part in attaining the goal of avoiding road and rail accidents and AI is at the forefront of some of the most promising innovations. For example, the privately-owned Brightline railroad has already proven itself to be the deadliest per-mile track in the US, in part because its locomotives run at up to 79 mph through densely-populated areas where the population is unused to higher-speed passenger rail; as a result, people regularly trespass onto the rails and many are tragically killed.
Given that Brightline is intent on expanding its line into Orlando and beyond – and that the costs of installing fencing along the trackside can exceed $200,000 per mile – another solution must be found. The company's decision-makers believe they have struck upon one in the shape of their collaboration with Remark Holdings, a tech and AI company whose Smart Safety Platform is capable of detecting intruders and identifying track anomalies from a distance, an innovation that should help to reduce accidents by a significant margin.
The road transport sector is enjoying a similar AI safety boost. While most media headlines focus on how AI can enable driverless vehicles, there is much lower-hanging fruit already being targeted by tech companies. For example, machine vision is capable of monitoring the health and performance of vehicle hardware, optimizing upkeep, and minimizing accidents caused by mechanical failure. So-called "cobots" (collaborative robots) can accelerate manufacturing processes, while AI, 5G, and thermal imaging technology can work in tandem to detect potential threats and share information between different vehicles.
What's more, traffic management has already benefited hugely from AI-based cameras located at intersections, with 155,000 expected to be in place by 2025. Meanwhile, Australian start-up Acusensus rolled out its Heads Up roadside camera network in 2019. Capable of identifying risky behaviors among drivers, the initiative saw an 80% reduction in the use of cell phones and a corresponding 22% drop-off in road accidents, winning prizes in the process. Given that the US recently passed the $1.2 Trillion Infrastructure Investment and Jobs Act (IIJA), the time is ripe for road safety to be revolutionized.
Although a world free from transport accidents might seem like a pipe dream, the forward march of technology could make it a feasible reality in the foreseeable future. Indeed, MIT research has even theorized that AI could leverage historical data to predict future events with a reasonable degree of accuracy, thus anticipating accidents before they happen and allowing users to take appropriate action to evade them. With such incredible opportunities within our grasp, it's time to bring AI fully into the fold and make transport-related injuries and deaths a thing of the past.
