Integration of Big Data Analytics in AV Inspection and Maintenance
11.3.1 Big Data Analytics' Role in Autonomous Vehicle Development
Although various technologies have provided small steps towards automation, a fully automated vehicle has yet to be realized. However, over the last decade, with the large range of advancements in technology and the newfound use of big data, tech companies have created the necessary programming for fully automated vehicles. AVs rely on the data they receive through a variety of systems, including the Global Position System (GPS), radar, sensor technology, and cameras.
These sources provide them with the data they need to make safe driving decisions. At this point, car manufacturers are still using stores of big data to work out the kinks of the thousands of scenarios an autonomous car could find itself in, but it’s only a matter of time before self-driving cars transform the automotive industry by making up the majority of cars on the road. The ability to use big data is changing industries worldwide, and one of the most essential big data analytics, deep learning, is contributing to the progress in the automotive industry towards fully AVs. Although it will still be several decades before the mass adoption of self-driving cars, the change will come. In only a few decades, we’ll likely be living in a time where cars are a safer form of transportation, and accidents are tragedies that are few and far between (Potter, 2019).
11.3.2 The Rise of Autonomous Vehicles
Technology adoption by the automotive sector has mainly centered on the concept of AVs and the efforts to make them a more viable form of transport. A UK government report estimates that the connected and AV industry will be worth £28bn in the United Kingdom by 2035, making this technology a prime focus area for all industry players. The industry is already capable of producing Levels 1-3 AVs (see Chapter 1), where higher levels correspond to reduced degrees of human intervention required during driving. The capabilities that make up these levels include environmental awareness, the real-time condition monitoring of the vehicle, and weather analysis. The ultimate goal is to enable AVs to reach Levels 4 and 5, where the driver’s input is minimal or the driver is made completely redundant. While this goal may only be possible in the distant future, autonomous technology has already begun to include features to make vehicles smarter and more capable of dealing with dynamic demands.
IoT and cloud integration are essential parts of the progression from “dumb” cars to smarter, intelligent vehicles that make better use of connectivity.
AVs today are integrated with sensors that constantly collect data in real time. These data are later stored in a remote cloud server for analysis. This effectively transforms the vehicle into a “black-box” recorder that can be used at any point to determine the current state of the vehicle’s functioning. The analyzed data are then used to create a customized driving profile which enables the in-built software to predict future driving scenarios and devise solutions to prevent accidents. In addition to in-car components, the integration of sensors into city environments for assets such as traffic lights, lamp posts, and traffic signs will help create a larger, more detailed database and improve situational awareness at all times (Singh, 2018).
11.3.3 Keeping Big Data Analytics under Control for Autonomous Vehicles
The driverless car has been a tech dream for decades, and now that broadband connectivity, cloud computing, and AI are increasingly available, autonomous cars may go mainstream, provided certain technical and regulatory milestones are reached.
First, Autonomous cars generate a staggering amount of data; Intel CEO Brian Krzanich estimates one car generates 4 terabytes of data in 8 h of operation. Multiple image, radar/laser-illuminated detection and ranging (LiDAR), time-of-flight, accelerometer, telemetry, and gyroscope sensors generate data streams that must be analyzed to perform the calculations, inspections, and adjustments required to safely navigate a car.
Second, analysis needs to happen in real time if the car is to keep up with constantly changing driving conditions (other cars or pedestrians moving around the vehicle, changing weather, traffic signs, etc.). These real-time performance requirements mean there’s no time to upload data to a central server, conduct the necessary analytics, and send instructions back to the car for execution. Thus, data that are critical to safely navigate the car must be analyzed locally by the car itself (essentially the car is an edge device in a cloud network).
Third, not only does the car need to analyze data on its own, it must also learn to pick and choose between different data streams to identify the ones best suited for analysis at any given moment to keep the car driving safely. This last requirement, the need to determine what data are required to perform an analysis, is tricky. While predefined filters can help a car’s machine learning routines learn what data to use and when to use them, those filters can’t be updated in real time. Accordingly, an autonomous car will need to run machine learning and analytics engines powerful enough to recognize mission-critical data requiring immediate analysis and action maintenance.
In other words, we need analytics and machine learning algorithms for autonomous cars that can:
- • Identify data in all formats.
- • Recognize what data are required for mission critical operations and analyze those data locally.
- • Compress or aggregate noncritical data for uploading to the cloud for future use.
- • Schedule uploads of noncritical data from the car to the cloud when less expensive communications are available (e.g., when the car is parked overnight at home and can access the owner’s Wi-Fi instead of a metered cellular network).
- • Know how to call for legacy data from the cloud so the AI can use it for future analytics.
The last one is particularly important. An autonomous car manufacturer will be responsible for storing vast amounts of data generated by cars operating around the world, and many of those data will likely have no real value when initially captured. However, their value may be revealed in the future as the manufacturer’s autonomous driving applications evolve and improve. Today’s noncritical data can be useful for future applications, provided the data are properly stored and catalogued so they can be easily found.
Without careful cataloguing of data as they are captured, autonomous car vendors run the risk of creating a “dark data” problem. Dark data is the term used to describe data an organization collects but fails to take advantage of because they don’t know how to or have forgotten they have them. This will be a significant problem for self-driving cars because of the sheer volume of data they generate. And as we see more vendors enter the autonomous driving market, the ones that will ultimately win out over others will be those vendors best prepared to analyze data at the local level and have catalogue their databases properly so future autonomous applications can find the legacy data they need, when they need them (Chala, Bayliss, & Camper. 2019).
11.3.4 Big Data Analytics—How Does It Help the Automotive Industry?
Continuous development of technologies is inevitably causing a transformation of the ecosystem of the automotive industry. The increasing demands from consumers coupled with diverse technological features, are making decision-making a tough process in the automotive industry.
However, with the availability of big data analytics, decision-making can be made easier (MARii, 2019). Ways big data analysis can help the automotive industry include the following (MARii, 2019):
- 1. Maintain smooth flow between links in the supply chain.
- • Auto manufacturers use huge amount of data from various systems, such as dealer management system (DMS), customer relationship management (CRM), and customer satisfaction surveys to gather metrics on numerous values ranging from customer inquiries to sales, inventory levels, customer order patterns on various models, trims, and color selection.
- • This process enables manufacturers to understand what their customers actually want in a particular vehicle.
- • In addition, the data collected allow suppliers to ensure the availability of materials needed to manufacture components that customers find appealing.
- 2. Enable strategic marketing planning.
- • The automotive industry is a platform for huge businesses.
- • Huge sums of money are invested in the automotive industry and it is likely to increase in value.
- • Thanks to the data collected through marketing mix analysis, comprehensive evaluation of customer responses, and internal business operations, big data analytics is helping auto manufacturers to strategically manage the use of incentives, rebates, financing deals, and other key attractions in increasing sales and return of investments (ROIs).
- 3. Make R&D easier.
- • The designing and engineering process of developing new vehicles involves a lot of decision-making dilemmas, often leading to the waste of time and money.
- - For example, a model in development may have a certain feature used in existing models, such as a door lock function button.
- - If the manufacturer leverages data monitoring and analytics from industry customer satisfaction studies, its dealership service departments, and its own internal research, concerns with that button based on existing models can be recognized.
- • This approach allows designers and engineers to avoid repeating the same mistakes, hence saving time and money.
- • The data sample can also enable the manufacturer to smooth the operational process; this allows more accurate material procurement, manpower planning, and vendor coordination.
- 4. Be data friendly and future-proof.
- • Data are tools to future-proof the automotive industry, significant enablers of the latest advances in the development of AVs.
- • The technology behind self-driving cars is machine learning or AI.
- • A machine “learns” to keep passengers safe through continuous research and development, facilitated by data collected from real-world driving conditions, environmental studies, and so on.
- - According to the Center for Automotive Research (CAR), one of the recent applications of big data analysis was the assessment of cost and effectiveness of powertrain technologies developed to meet global standards in fuel economy and greenhouse gas emissions.
- - The research found a discrepancy between what automakers and regulators project as the cost of meeting those standards.
- • This shows that data also have the ability to inform regulators and policy makers as they make decisions going forward, and this could benefit automakers in the long run.