The role of deep data and intelligent edge filtering in the future of connected vehicles

Traditional automotive data analysis methods involve moving all on-board data to the cloud for processing. This is accomplished either with an over-the-air (OTA) software solution, which is often limited to infotainment and navigation data, or at the dealership with a direct data download.

When working with a limited number of data points or parameters, say ten for this example, this method is not great. But what happens when this number increases tenfold? A hundredfold? The truth is that it will soon be much more.

By 2025, it is estimated that connected vehicles will produce data at a volume of 10 exabytes (10 million terabytes) per month. That’s more than 1,000 times its current volume. This is ideal for OEMs, who can use this data to create downstream revenue, improve component efficiency and functionality, and reduce recall and warranty costs. However, a major problem stands in the way: the ability to access and process this data at scale.

Fortunately, the Deep Connected Platform (DCP), an IoT-based embedded firmware and cloud solution from Sibros, opens the door not only to deep data but also to real data. This type of vertically integrated software and data management solution enables OEMs with multiple communication methods (CAN, LIN, Ethernet, Flex) to operate embedded systems and provides millisecond-level data collection and intelligent edge filtering from pre-development to product deployment and beyond. Let’s dive deeper into the backbone of connected vehicle software and data access.

Interpret deep data with intelligent edge filteringMost of today’s automotive data information is gathered from a lab simulation. In other words, component improvements and decisions are made based on the “best guess” rather than accurate, real data. Unfortunately, there is no way of knowing how close these virtual tests are to real life or if the decisions made will prove effective for the functionality and performance of the vehicle.

Sibros DCP opens the door to real-world data, not just from one or two vehicles, but from the entire fleet. Its superior data compression methods and intelligent edge filtering speed up the process of moving data from vehicle to cloud. In other words, DCP brings data sorting, processing, and compression closer to the data source and enables data separation for smoother analysis. Customizable settings allow OEMs to isolate information for a specific system, such as Advanced Driver Assistance Systems (ADAS) or EV Battery Management Systems (BMS). Rather than digging through a pile of irrelevant details and data points, manufacturers can use selective data logging and customizable frequencies to examine exactly what they want, when they want to quickly get actionable results. .

Vehicle predictions
When the oil light comes on, it indicates a need for preventive maintenance: change the oil now to increase the longevity of your vehicle later. The problem is that oil change frequency, whether determined by the car or by a mechanic, has always been determined with two data points: time and distance.

The problem with this type of reasoning is that additional factors contribute to engine oil requirements. Driver habits, terrain, outside temperatures, and even undetected system faults all contribute to engine oil cycling. Sibros Deep Logger is responsible for collecting data from every Electronic Control Unit (ECU) in the vehicle. With this information, manufacturers can consider system dependencies and driver habits to not only provide more accurate preventative maintenance recommendations for oil changes, but for other systems as well. Preventative maintenance is just the tip of the iceberg. Another, and perhaps more important, benefit of in-depth data and intelligent edge filtering is the ability to identify and correct software and hardware faults that compromise vehicle and owner safety.

Improve data with OTA updates
According to the Society of Indian Automobile Manufacturers (SIAM), there were over 7 lakh vehicle recalls in 2021, a number that will only increase along with the complexity of the vehicles. Unfortunately, many recalls go undetected for months or even years. Then there’s the issue of completion rates, which range between 56 and 76 percent and over time result in millions of unsafe vehicles on the road. These issues are easily resolved with in-depth data access, analysis, and live updates.

Let’s say the brake assist in the vehicle’s ADAS experiences a 0.05 second delay after detecting a potential collision. Without access to real-time data, it may take several traffic accidents or near-misses before this problem is revealed. A simpler solution is to identify and fix the problem before it escalates. With Sibros DCP, OEMs can set up alerts that notify them when a particular ECU is operating outside of acceptable parameters. This is especially crucial with safety-critical components where faults can quickly escalate into life-threatening issues.

The next step is to determine the scope. Is the fault limited to a specific make, model or transmission? Is this an isolated incident? Or maybe it’s a fleet-wide design flaw. Gone are the days when OEMs had to call back every suspect unit to figure this out. They can accomplish this remotely with millisecond level data extracted by Sibros Deep Logger directly from the vehicles ECUs.

Sibros further streamlines the recall deployment process by enabling vehicle-wide over-the-air (OTA) software updates with its Deep Updater. Instead of waiting for customers to respond to a recall summons, OEMs can increase completion rates with a targeted update package sent via 5G or Wi-Fi directly to affected vehicles. Once the package has arrived, it waits for the owner’s agreement before starting the installation. Safety-critical updates can be sent with a time constraint: if owner consent is not received within a specified time, the update will initiate the next time the vehicle is in a safe state . Not only does this keep dangerous vehicles off the road, it improves the driver experience by eliminating a visit to the dealership.

Customer-oriented features

Deep data and intelligent edge filtering also play a key role in creating innovative and desirable features for customers. It allows OEMs to see what systems and features pilots use and how they use them.

The next step is to take this information and turn it into revenue streams, for example by providing the ability to remotely upgrade in-vehicle functionality. Some examples of this are on-demand engine calibrations for different types of terrain, automotive windshield wiper adjustments for heavy rain, automatic window defrosting, and heater settings for ice and sludge removal. fogging, and custom cockpit settings to increase driver comfort. In addition to direct upgrades, other customer-facing features include access to custom third-party apps and more accurate usage-based insurance pricing.

Stay on the cutting edge with Sibros
Deep data and intelligent edge filtering are key players in the evolution of the connected vehicle and are easily accessible with a hardware-independent platform, such as Sibros DCP. The Deep Connected platform can be used both as an out-of-the-box solution or as an add-on to enhance a manufacturer’s existing automotive software. Sibros complies with global security, cybersecurity, and data management standards to enable seamless solution integration and deployment.

DCP offers compatibility with all vehicle architectures, from ICEs to EVs and two-wheelers to passenger cars to provide OEMs with scalable fleet-wide updates and data management. Sibros enables a safer driving experience with safe, secure and robust OTA software updates for the entire vehicle lifecycle and helps OEMs reduce transfer and cloud storage costs with a superior data compression at the edge. Don’t get stuck in the past with limited data access and sorting capabilities. Drive into the future with Sibros Deep Connected Platform, a safe, efficient and reliable way to manage data and power your connected vehicle ecosystem.

Comments are closed.