Uzi Baruch, general manager of Electronics at OptimalPlus says autonomous vehicles have already arrived and urgently need lifecycle analytics for their safety now and in future. And no, we’re not just talking about prototypes from the likes of Google, Uber, and Tesla.
Autonomous vehicle technology encompasses a wide range of tools and features with varying levels of sophistication. The fully self-driving car may represent the apex of autonomous vehicle development, but autonomous technology is embedded in every new car that’s manufactured today (nearly all are level 2 to 3), from driver assistance systems like cruise control to mechanisms that alert drivers to hazardous road conditions.
The level of advancement now is built on thousands of electronic components, a spectacular leap from a decade or so ago, when electronic components in cars were a fraction of what they are today. Further complexity is added by the fact that many of these technologies went from concept to working part in a matter of months, instead of being the tried and tested technology that was in the cars of yesteryear.
Disparity and IP worries
These thousands of sophisticated parts are produced by disparate factories under different ownership, who, fearing intellectual property theft, are wary of sharing data with each other, resulting in heavily siloed critical information that is difficult to analsye. Add to the mixture the massive number of data points in the manufacturing process, which is itself becoming much more sophisticated, and established analysts (manual or human) simply can’t process and analyse this abundance of information efficiently.
In short, it’s a complex process with few signs of slowing down. Compare the clunky, boxy cellphones that were commonplace just a decade or two ago with the modern, streamlined version of today. The process didn’t happen overnight – and manufacturers paid the price for trying to speed things up too quickly, without making sure that the electronics were reliable (see the exploding Galaxy note disaster.)
Starting up with funds
The growing number of start-ups dedicated to improving autonomous vision are receiving incredible amounts of funding. They are part and parcel of autonomous car development and deployed in countless variations today. However, by focusing on the advancement of autonomous vehicle vision alone, many are blind to for the hidden element necessary for ensuring the safety of today’s autonomous vehicles and ensuring the safety, development and trust that are necessary for the autonomous vehicles (level 4 to 5) in future – lifecycle analytics.
Lifecycle analytics is a holistic, comprehensive solution for detecting manufacturing issues in real time. Using big data that encompasses the full cycle of manufacturing data, manufacturers are empowered to take decisions based on data-grounded approaches.
Performing deep analysis of manufacturing data, mechanical operations and advanced visual inspection using sophisticated Image processing algorithms, lifecycle analytics is the only way to predict and detect potential product issues on the factory floor.
Lifecycle analytics is not just essential for consumer and manufacturer protection – it is a catalyst for the autonomous revolution.
Electronic components, including those present in every advanced driver assistance system, have gone from important to mission critical. The failure of an electronic component doesn’t just mean returning the car to the manufacturer, it could mean that you stray out of your lane, or far worse. Yet, despite these advances in technology, implementation of systems that can ensure their reliability and dependability has been left behind far too often: there were three times more recalls due to electronics across the industry in 2016 than there were in 2014.
In 2017, 400,000 Jeep, Chrysler and Dodge cars recalled in 2017 due to electrical component failure, a shocking statistic that could rise. It goes without saying that the introduction of autonomous vehicles and fractional ownership, which could see cars used up to 22 hours a day, also will require a much more intensive manufacturing process, with all the testing and additional electronic components that will entail.
The big picture is key
When it comes to guaranteeing vehicle reliability and safety, it’s easy to get bogged down in the details of thousands of tiny electronic components and mechanical operations. Lifecycle analytics gives manufactures the big picture based on millions of parameters in the entire process of car manufacturing, revealing the entire picture in a way that empowers manufacturers with invaluable actionable insights using big data technologies and a data centric approach. Without this overview, truly ensuring dependability is impossible.
By deploying holistic, long-term analytical solutions for the supply chain, car manufacturers can expect fewer recalls and more reliable vehicles. Industry experts estimate that auto recalls cost manufacturers a staggering $22.1 billion (€19.4 billion) in 2017 alone, so that means a vast amount of money saved for manufacturers today – not to mention avoided collisions and road fatalities. It’s a win for consumers and manufacturers, but it’s equally, if not more vital, for laying the groundwork for progressing our autonomous future. Consumers’ trust slips with every crash, delaying the implementation of new developments and heightening government scrutiny.
No matter the blistering pace of self-driving technology, the (supply) chain is only as strong as its weakest link. That fact, along with the increasingly critical nature of electronic components in cars today, makes lifecycle analytics the most important advancement out there for autonomous vehicles. If companies really want to provide the safety consumer deserve and the advances they demand, implementing lifecycle analytics is a must.
For more information see Optimal Plus
The author of this blog is Uzi Baruch, general manager of Electronics at OptimalPlus