As sensor prices fall, the technology will become increasingly central to automotive manufacturing, says intelligence firm CB Insights, allowing car makers to monitor automated and manual assembly lines, streamline their operations, and predict costly outages before they occur.
The analysts’ latest report on Auto and Mobility Trends – mobility in the sense of connected and autonomous transport – provides a comprehensive overview of tech hotspots and adoption cycles across the sector.
For example, Audi has built a smart factory in Mexico, where it uses radio frequency identification (RFID) tags and smart logistics to coordinate entire production runs. The company has also deployed RFID at its Neckarsulm factory in Germany to identify vehicles at every stage of the manufacturing process.
Sensors also allow automakers to track components through the value chain and employ predictive analytics. CB Insights quotes the example of BMW’s Connected Supply Chain (CSC). This enables the German car giant (which owns the Mini and Rolls Royce marques) to respond more quickly to delays, and so minimise the costs associated with production slowdowns.
But what about sensors within connected and autonomous cars themselves?
LiDAR (light detection and ranging) allows vehicles to ‘see’ accurate 3D visualisations of their surroundings, formed by sending out pulses of light. Despite being dismissed by some manufacturers, LiDAR is fast emerging as a more viable and affordable technology than before, suggests the report.
Until recently, some manufacturers have charged over $100,000 for LiDAR units, making them impractical to deploy at scale or outside of dedicated test vehicles. Traditionally, these systems have also relied on fast-spinning parts, rendering them both expensive to produce and less reliable than static devices.
However, a number of startups are now working to improve the cost and performance of these sensors – notably with solid-state LiDAR, which has no moving parts, explains CB Insights.
According to the report, major automakers and Tier-1 suppliers such as BMW and Aptiv have partnered with – and are investing in – solid-state LiDAR startups in recent months. For example, BMW announced a deal with Innoviz last April, which is also working with Magna.
Despite this progress, LiDAR still has limitations other than cost, suggests the report. Many systems can be affected by poor weather, such as heavy rain or fog, which can obstruct the light pulses. As a result, the industry consensus on sensor technologies is to use LiDAR as a complement to cameras and radar, rather than risk creating a single point of failure.
A vehicle could then combine data from a variety of different sensor types via fusion technology or dedicated sensor-data analysis software, suggesting that these will be growth markets alongside the sensors themselves.
With two fatal accidents last year involving cars running under software control, the need to create solutions that can deal with the unpredictable human world – in all weathers – is essential, if autonomous cars are genuinely to be safer than human drivers. It’s also vital to improve public trust in driverless systems, which has fallen since the Uber and Tesla crashes last March.