Upstream: Better Data Collection, Processing Will Reduce SDV Recalls

October 1, 2025

Sebastian Blanco

The 2025 Bronco Sport, seen here in Heritage Edition trim, is just one model Ford had to recall in 2025 as part of a historic 100+ recalls in the first eight months of the year. (Ford)

Cars that are more connected, equipped with more sensors than ever before, should make proactive maintenance somewhat easy and reliable. Drivers could have lower repair costs and fewer breakdowns overall if the automotive industry shifts away from a periodic maintenance mindset towards data-driven proactive services. But the automakers themselves would also win with a massive drop in recalls.

A connected car requires accurate telemetry sensors to track details such as temperature (in the engine, battery, and cabin), pressure (in tires, fuel, and oil), electrical current/voltage, and vibration patterns to detect problems before they become failures. The vehicle also needs to be able to combine telemetry, diagnostics, metadata, and service records in one platform and then make sense of it all. “Poor data integration kills even the best analytics,” according to Upstream co-founder and CTO Yonatan Appel.

Upstream, a provider of AI-powered cloud-based data management platform for automotive, smart mobility, and IoT ecosystems, released a report in July that looked at how connected vehicle data was and could be used in “next-generation after-sales quality strategies.” Upstream advised OEMs to use connected car data not just to provide better customer service, but also to “better leverage data-driven detection and analysis to reduce the risk of high-impact recalls, warranty costs, and customer dissatisfaction,” especially when it comes to EVs recalled for software issues.

The underlying problem, Appel told SAE Media, is that preventative tools in connected cars are not being used to their fullest extent. Upstream claimed in its report that 70% of all the recalls it studied – and nearly 90% of EV-related recalls – could have been detected earlier using connected vehicle data and AI. Appel agreed with SAE Media that those are pretty dramatic numbers.

“They are,” he said. “We were surprised, but the AI-enriched data told the story.”

Upstream singled out Ford for setting what Upstream called “an unfortunate industry record: 88 safety recall events in just the first half of 2025 - more than any automaker has ever reported in a full calendar year.” Upstream pointed out that Ford might be the most visible example of the growing recall trend – since July, Ford has had to announce more recalls, and the number now stands at over 100 so far this year, affecting more than two million vehicles - but it’s not the only OEM facing potential problems.

Upstream analyzed structured data from NHTSA’s recall and complaints databases for its report, limiting the data to issues that occurred in 2020 or later and that were associated with vehicle models from model year 2020 onward. Upstream removed records related to child car seats and tires from the dataset, which went through April 2025 for recall data and the end of 2024 for consumer complaint data. This left Upstream with a report that included insights from more than 5,000 recalls and 30,000 consumer complaints, as published by NHTSA. To prevent skewing the results with cross-model repetition, Upstream treated each recall campaign as a single record, with duplications across different affected vehicle models removed. Then, Upstream applied AI tools to enrich and classify the data.

“These processes enabled us to identify trends and categorize issues by various factors, like OEM vs. supplier, vehicle engine types such as EV and ICE, and related data signals,” Appel said.

For its part, Ford said it is aware of software problems.

“We are making progress on software quality, using an enhanced software validation process to help ensure the right software is present on vehicles and using over-the-air updates to address potential warranty issues before they become customer issues,” a spokesperson told SAE Media. “Modern digital safety systems require new talent, tools, and standards. Ford has deeply invested in those capabilities like connected data signaling and dramatically increased testing on all new products across broad variations in customer use. We use multiple sources to detect quality issues, including social media, repair orders and connected vehicle data. Where possible, we’ll use over-the-air updates to repair software-related recall issues.

EVs, and SDV technology in general, offer both risk and the opportunity for early problem detection, Upstream said. Since EV platforms rely on software-defined, data-driven systems, their chance of succumbing to a large-scale recall is higher, even though that doesn’t have to be true. “EVs generate a continuous stream of operational insights that can be used for quality assurance: nearly half (49%) of EV-related recalls could have been identified early through diagnostic trouble code (DTC) monitoring alone, compared to 37% for all recalls,” the company said.

“The bottom [line] is to get clean data, provide proper context, and keep engineers in the loop,” Appel said. “This approach catches quality issues quickly and helps to speed up the investigation process.”