Porsche Is Using AI To Catch EV Battery Problems Before They Happen
I'm tired of AI getting shoved into every corner of life in 2025. But, the technology does have uses in cars beyond gimmicks like fake conversational personas. Like sifting through data and identifying trends to optimize an EV battery pack, for example. That's one of the things Porsche is using AI for, and the automaker recently pulled back the curtain on how it's implementing machine learning to make its electric cars better.
In a press release from earlier this month, Porsche focuses on two particular aspects of the way it's using AI to monitor battery health. First, there's the component of predictive quality assurance, which boils down to making informed insights into how the battery will age, based on data from across the company's user base. But arguably more valuable than that is the system's ability to detect anomalies, which can stop a potentially serious battery defect before it starts.
Porsche doesn't explicitly mention thermal runaway in this release, but that's where my head goes as I'm reading this. In a thermal runaway, a problem with one cell quickly propagates to others. But there may be signs beforehand that something's not right, and that's why it's especially valuable that Porsche says it can monitor battery health on a cellular level and proactively warn customers, with instructions relayed through the owners' app. Porsche recently set a record at Road Atlanta for the fastest lap for a series-production electric car with the Taycan Turbo GT. Porsche
Like anything related to safety in the automotive realm, you hope you never have to encounter a situation like this with your vehicle. But knowing these safeguards are in place certainly offers peace of mind. And this kind of deep learning can be leveraged in other areas, too. State of charge and charging speed, for example, should be intelligently managed for any device that incorporates a lithium-ion battery, from your smartphone to your car, to extend the pack's lifespan as long as possible.
It's encouraging that Porsche seems to be using AI not to develop an in-car chatbot that will pretend to care about your problems—or something else nobody's ever asked for—but to ensure the long-term performance and safety of its products. They sum it up well in this excerpt: 'For Porsche, AI is a tool that helps the team to understand complex relationships and take all relevant aspects into account.' Using new tech to better nail the fundamentals is the kind of future we can get behind.
Got tips? Send 'em our way at tips@thedrive.com.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
22 minutes ago
- Yahoo
Analysis-Europe's old power plants to get digital makeover driven by AI boom
By Forrest Crellin PARIS (Reuters) -Some of Europe's ageing coal and gas fired power plants can look forward to a more high-tech future as big tech players, such as Microsoft and Amazon, seek to repurpose them as data centres, with ready-made access to power and water. Companies such as France's Engie, Germany's RWE, and Italy's Enel are looking to benefit from a surge in AI-driven energy demand by converting old power sites into data centres and securing lucrative long-term power supply deals with their operators. The data centre option offers the utilities a way to offset the hefty costs of shutting down ageing power plants as well as potentially underwriting future renewable developments. Tech companies see these sites as a quick way to secure power grid connections and water cooling facilities, two big bottlenecks in the AI industry. "You have all the pieces that come together like ... water infrastructure and heat recovery," said Bobby Hollis, vice president for energy at Microsoft. Lindsay McQuade, EMEA energy director at Amazon , said she expected permitting for data centres to move faster at old sites, where a big chunk of infrastructure was already in place. Utilities can either lease the land or build and operate the centres themselves, securing long-term power contracts with tech firms, he said. The deals offer much more than just the sale of unused land as they include opportunities for stable, high-margin revenue, said Simon Stanton, head of Global Partnerships and Transactions at RWE. "It's more about the long-term relationship, the business relationship that you get over time that enables you to de-risk and underwrite your infrastructure investments," Stanton said. Most of EU's and Britain's 153 hard coal and lignite plants are set to close by 2038 to meet climate targets, joining the 190 plants that have closed since 2005, based on data from NGO Beyond Fossil Fuels, which campaigns to accelerate closure of coal-fired power stations. NEW REVENUE STREAMS The economics of data centre deals can be compelling for the utilities, which can negotiate a long-term power supply contract to underwrite future renewable developments. Tech firms are paying premiums of up to 20 euros per megawatt-hour for low-carbon power, said Gregory LeBourg, environmental program director at French data centre operator OVH. Data centre power demands can be anywhere from a couple hundred megawatts to a gigawatt or more. So the annual 'green premium' - the extra price paid for low-carbon electricity - on top of a base market price could potentially translate into a long-term contract worth hundreds of millions or even billions of euros, based on Reuters' calculations. One long-term option is to build an "energy park" and connect the data centre to a new renewable development, relying on the grid for emergencies, but this is a relatively new concept, industry sources said. Engie wants to double its installed renewable energy by 2030 from the current 46 GW. The group has identified 40 sites globally that it is marketing to data centre developers, including coal and gas plants that could be converted, said Sebastien Arbola, who runs the company's data centre business. One is the Hazelwood coal plant in Australia, which closed in 2017. He declined to disclose details of other sites, saying they are mostly in Europe. Other utilities, including Portugal's EDP, EDF, and Enel said they are also marketing old gas and coal sites for new data centre development. "It's business model diversification," said Michael Kruse, managing partner at consultancy Arthur D. Little. Utilities are creating a new type of business and also new revenue streams, he said. 'SPEED TO POWER' The appeal for tech companies is speed. Grid connection delays in Europe can stretch over a decade, while repurposed plants potentially offer speedier access to power and water. "You actually have the opportunity to move faster," said Hollis at Microsoft. Data centre capacity in Europe is much lower than the United States and Asia due to longer grid connection times and slower permitting, data from Synergy Research Group showed. The data centre operators can choose to buy the renewable power they need directly from the utilities in the form of long-term contracts or purchase from the power market. Real estate firm JLL is working on several conversions, including a 2.5 GW data centre at a former German coal plant and four sites in Britain for a major tech client, said Tom Glover, who works on data centre transactions at JLL. Developers do not often disclose more detail about data centre projects, including their clients, for security reasons. Britain's Drax is also seeking a partner to develop unused parts of an old coal site in Yorkshire, now partially converted to biomass. It offers access to unused water cooling equipment, said Richard Gwilliam, Drax's carbon programme director. Drax is offering a "behind-the-meter" deal where the power plant will provide direct power to the data centre and it can pull from the grid if necessary. EDF has also chosen developers for two sites at gas power plants in central and eastern France. Tech companies are willing to pay more for projects that can start up sooner as they vie for market share in a rapidly growing industry, said Sam Huntington, director of research at S&P Global Commodity Insights. "Speed to power is just the phrase we keep hearing over and over again," he said. Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
Forbes
25 minutes ago
- Forbes
Scaling Production AI: 20 Surprising Hurdles And How To Overcome Them
Scaling AI from prototype to production is a major hurdle for many organizations. Even with skilled teams and powerful models, companies often hit unexpected bottlenecks—such as disjointed data pipelines, unclear ownership or low user trust—that slow or stall progress. To scale AI successfully, businesses need more than just technical expertise; they must invest in robust infrastructure, change management and cross-functional alignment. Below, members of Forbes Technology Council reveal common hurdles teams face when scaling production AI—and share practical strategies for overcoming them. 1. Instilling Digital Literacy And Trust Among Team Members A major bottleneck in scaling AI is organizational resistance due to fear and digital skill gaps. Many employees worry that AI will replace their roles or feel unprepared to use it. This often leads to low engagement and stalled adoption. To overcome this, companies must invest in AI literacy and upskilling. Involving staff early in the AI journey fosters trust, making scaling smoother. - Praveen Tomar, UK Civil Services (Ofgem) 2. Choosing The Right Problems To Tackle AI holds promise in enterprise restaurants—from predicting inventory to personalizing upsells. However, a common bottleneck isn't technical; it's misalignment between what's built and what drives value. Too often, teams chase flashy use cases without validating impact. Scaling AI starts with choosing the right problems and having the right data. - Peter Kellis, TRAY Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify? 3. Providing Frontline Access One key bottleneck is providing direct access for the frontline workers who would benefit from it. Providing frontline workers with AI solutions to efficiently and effectively complete their tasks is critical in today's landscape, where time is of the essence. - Mike Zachman, Zebra Technologies Corporation 4. Building Trustworthy Data Architecture Fragmented data architectures fragment business context. Without integrated enterprise knowledge, AI lacks organizational understanding. Success requires a trusted data architecture that's operationally simple yet preserves business meaning. - Louis Landry, Teradata 5. Avoiding Prompt Injection And Tool Misuse Prompt injection and tool misuse are critical bottlenecks. Attackers can trick LLMs into leaking sensitive data, performing unauthorized actions or amplifying these to widen impact. This slows deployment and can lead to costly breaches. You can mitigate risk with the same security best practices that you use elsewhere in the organization: least privilege, sandboxing, audits and adversarial testing. - Willem Delbare, Aikido 6. Managing AI-Generated Content Overload AI will generate more content than existing review and approval processes are designed to handle. Companies will need to decide how to proceed—they can either accept the risk, dramatically slowing adoption, or find and/or build tools to help them manage it. - Larry Bradley, SolasAI 7. Identifying And Managing ML Pipeline Cost And Performance Issues Companies often overlook MLOps when they take on AI-related development. ML pipelines can have very different cost and performance challenges than traditional development domains. Understanding the true unit cost of model performance is essential and can mean the difference between a model that scales well and one that doesn't. - Siamak Baharloo, Labviva 8. Linking AI Spend To Business Value One big bottleneck? The CFO calling out the inference cost line on the P&L. If teams can't clearly link AI costs to business value, there's a real risk that projects will stall. The solution? An AI Command Room. Start by building a clear business case tied to at least one normalized metric, such as labor dollars saved through AI agents and automation. Then, track and report impact—from hypothesis to daily execution. - Matt Kesby, Multiplai Tech 9. Establishing Shared Data Definitions As enterprises deploy agentic AI, agents make decisions based on data no one fully trusts. In pilots, humans catch issues. At scale, autonomous systems pull from disconnected sources with conflicting rules, and small cracks become major failures. The issue isn't the model; it's the lack of data trust. Scaling AI requires shared definitions and automated checks to ensure a reliable foundation. - Jay Limburn, Ataccama 10. Grounding Data In Real-Time, Trusted Context A key bottleneck in scaling AI is context fragmentation—models can't reason effectively without access to clean, connected and secure organizational data. Grounding data in real-time, trusted context is the real challenge. The fix is better infrastructure: unified knowledge layers, privacy-preserving compute and retrieval pipelines that align AI with the right data. - Regan Peng, PINAI 11. Addressing Data Lineage And Quality Debt One unexpected bottleneck is data lineage and quality debt. Many organizations assume that once a model is trained and performs well in testing, scaling it into production is mostly an engineering and computing problem. In reality, the biggest bottleneck often emerges from inconsistent, incomplete or undocumented data pipelines—especially when legacy systems or siloed departments are involved. - Sandeep Uthra, OneAZ Credit Union 12. Working With Outdated Infrastructure A bottleneck would be inadequate data infrastructure, which can be overcome by investing in modern, scalable data platforms and robust engineering practices. Companies should be establishing robust data governance practices and centralized data infrastructure early in the development lifecycle. - Nazih Chamtie, KMicro Tech, Inc. 13. Managing Model Drift Even though data quality is an obvious bottleneck, when we scaled AI in production, I primarily expected challenges with data and infrastructure. However, what caught us off guard was model drift. For example, in finance, market shifts rendered our fraud models stale. In healthcare, evolving patient data skewed predictions. Real-time monitoring and retraining turned out to be just as critical as building the models. - Dr. Suresh Rajappa, KPMG LLP 14. Overcoming Resistance To AI Integration Change management within the existing user experience is a significant challenge when scaling AI. While AI has enormous potential to enhance user experience, effective integration requires more than technical implementation. Organizations can proactively address this issue by increasing user adoption, reducing resistance and unlocking AI's full transformative potential. - Shivaprakash S Nagaraj, Digit7 15. Struggling To Find Skilled Architects And Operators Designing and building infrastructures that support AI initiatives requires a different set of skills. While many current IT professionals will get up to speed over time, there are significant nuances that come with AI that they may have never experienced before. Struggling to find skilled infrastructure architects and operators who already understand these nuances will be the reality for a few years. - Mike Wong, Accton Technology 16. Developing A Single Source Of Truth AI can be hindered by a lack of data access, data silos, data quality issues, missing context and so on. A lack of a single source of truth is also a common issue—especially if data connections are intermittent and complex—and AI models need to be run locally. Thinking strategically about the desired business outcomes and implementing a data fabric will improve data management and help scale AI end-to-end. - Heiko Claussen, Aspen Technology, Inc. 17. Identifying And Patching Security Gaps Security is the most significant challenge, because LLMs can't be controlled. Even foundational model providers' system prompts and training instructions have leaked. Strengthening AI security requires identifying potential vulnerabilities and testing against them, but we do not yet have an exhaustive list of those vulnerabilities, as AI systems remain a black box. - Rohit Kapoor, Tekmonks 18. Connecting AI Outputs To Decision-Making Processes A major bottleneck is the lack of integration between AI models and core business workflows. Even the best models fail to deliver value if their outputs don't connect to decision-making tools or processes. You can overcome this by embedding AI via APIs, aligning with existing systems and designing with end users to ensure seamless adoption and continuous feedback loops. - Motasem El Bawab, N3XT Sports 19. Detecting Performance Degradation Companies can build decent AI models but struggle to scale them due to poor production monitoring and deployment capabilities. Unlike deterministic code, where 2 + 2 always equals 4, AI models produce variable outputs, making it hard to detect performance degradation. Companies can overcome this by implementing robust MLOps pipelines with continuous monitoring, A/B testing, and automated alerts for model drift detection. - Mia Millette, Skyline Technology Solutions 20. Ensuring Organizational Alignment Accelerating AI initiatives without addressing data quality, readiness and strategic alignment will undoubtedly result in significant roadblocks. To scale successfully, organizations must prioritize use cases based on business impact, ensure data is properly prepared, and tightly align each initiative with clear, measurable outcomes. - Sean Nathaniel, DryvIQ
Business of Fashion
an hour ago
- Business of Fashion
Unpacking Fashion's New AI Marketing Toolkit
Ever since Mango started to use Smartly, an AI-driven performance marketing tool, for social media ads, its revenue generated by these ads has quadrupled. Using the platform, Mango's design team is able create brand templates that can be easily toggled between different versions of a photo of a dress, for example. Whereas before, creating different images required taking photos of different outfits in multiple settings. The womenswear brand is far from the only fashion player to have figured out how to use artificial intelligence to optimise and boost the efficiency of advertising. Companies including sneaker retailer Foot Locker and German e-commerce giant Zalando are using AI for everything from product image backgrounds to quick-turnaround campaigns that align with social media microtrends. Consumers go through content — and trends — so quickly today that advertisers struggle to keep up with analog creative production alone. AI's rapid generation capabilities help marketers produce more varied ads so viewers are less likely to tire of them as they scroll social media's endless feeds. But figuring out the right formula to use AI is no small feat. New tools seem to pop up weekly and while some platforms excel at automatically generating images, they may lack the editing tools marketers need to ensure visuals stay on-brand. Brands must consider their unique needs — which can range from visualising creative ideas for a campaign and enhancing their product imagery with dynamic backgrounds to automatically targeting different audiences with different versions of the same ad — in order to determine the ways in which AI could amplify their creative production. The most accessible AI tools are ChatGPT and Adobe's Firefly, which many agencies already use for easy image generation. Others favour more advanced platforms like Midjourney and Leonardo AI for greater creative control and experimentation. Meanwhile, AI agency Maison Meta and Zalando aggregate multiple AI technologies into their own platforms, curating the best tools for specific tasks like background generation and editing under one interface — Maison Meta also hosts workshops for clients like Mango to teach them how to use its toolkit. More advanced tools are able to connect the dots between image generation and ad optimisation on social media. Meta, for instance, has its own suite of AI tools called Advantage+, which includes creative enhancements that allow brands to customise their ads by animating imagery, adding music or enhancing copy. Platforms like Smartly and Pixis are also dedicated to optimising the creative content in ads, allowing marketers to create visual templates that can be automatically updated to cater to different customers across platforms like Meta, TikTok, Amazon and more. Footwear retailer JustFab, for example, was able to create a template for its broad product catalog by removing the backgrounds across all images and generating new seasonal scenes across using Smartly. 'It's the consumption happening now through social media that requires a volume that the marketing industry wasn't prepared for,' said PJ Pereira, creative chairman of advertising agency Pereira O'Dell. Lead With Human Creativity Regardless of the AI tools, human creativity and artistic direction must come first. 'This is not the end of the creative director or the creative teams,' said Jason Widup, senior vice president of marketing at AI advertising company Pixis. 'What we're seeing is the creative teams want to create new, interesting, fun concepts. They want to be pushing the envelope. And AI still doesn't have human taste.' What this looks like in practice is using traditional brainpower to come up with core campaign concepts, and then using AI to see how far the idea can be pushed — testing out how an image would look on Mars, for instance, or with a giraffe alongside a model, said Pereira. 'You can think and see in real time,' he said. While in the past, campaign generation could be costly, AI enables marketers to test multiple ideas at one time and see what works best. 'The biggest leap currently as a creative industry is the fact that the risk of trying something is infinitely smaller,' said Pereira. Turn One Asset into Countless Variations Proponents of the technology still don't trust AI to fully take over ad production, said Pixis's Widup. Instead, many feed product assets and brand guidelines into tools like Smartly and Pixis, and then use the platforms to iterate on existing ideas and imagery to change backgrounds, colours or placement to create different versions of an ad that can be personalised and targeted across audiences. For Zalando, using its own AI tools to showcase products in a relevant context — like a hiking boot on a mountain, or even to turn static imagery into video — makes all the difference in driving customer engagement. This, in turn, helps them better understand what their customers want to see, which leads to fewer returns and triple the conversions. 'We do not have endless hands, and we do not have endless time and endless budget,' said Matthias Haase, vice president of content solutions at Zalando. Localise in Real Time Beyond creative content, AI can be leveraged to reach the right users at the right time. Whereas in the analog days of digital acquisition, consumers were repeatedly served the same ads, brands can now automate when viewers are served a piece of content — whether for a complementary product to one they already own or a specific handbag colour they've been searching for. Most of Zalando's AI-generated content is market-specific to cater to local moments like Oktoberfest in Germany or regional running events, for example. These quick-turnaround, targeted campaigns typically take four days to produce, compared to six to eight weeks for traditional ad campaigns, and made up 40 percent of Zalando's campaigns in 2025. Its product description pages will eventually be customised to each geographical market using AI, showing different outfit combinations for French versus Polish shoppers, for example, and swapping out models that are recognised in different regions using digital twins. Build Content that Responds to Context For platforms that connect AI image generation with targeting such as Smartly, brands can reach customers even more dynamically by making each element of an image changeable depending on the customer who is interfacing with it. 'Every element can effectively change based on the time of day, [and the] specific audience that is being targeted,' said Oliver Marlow-Thomas, chief innovation officer at Smartly. 'We can dynamically map specific audiences to show specific variants … The dress the model is wearing, the shoe that's in the image, the colour of the background, or the colour of the hat, all of that is mapped and changeable.' Pixis, on its end, draws on data sources from Shopify to Google to macroeconomic data and weather channels to inform its decisions around which ad imagery is shown to who and when, said Widup. The platform has tested these dynamic ads for fashion brands on Meta, tailoring visuals and messaging to current weather patterns to make them feel relevant — showing rain jackets during storms or breezy outfits during heatwaves, for example — and boosting engagement and conversion rates. Optimise Targeting with Platform Data Brands that advertise on Meta can use its native AI tools to enhance existing content and optimise targeting. After Meta's targeting capabilities were reduced with Apple iOS changes in 2021, the platform began to build out Advantage+ in order to enhance advertising effectiveness on the platform by testing different ad types across users — and the more versions of creative content it has to test on customers, the better. 'Platforms like Meta are now saying, 'We've reduced your targeting capabilities, but just give us a lot of creative ads to use, and we'll refine it,'' said Widup. Meta's own targeting only goes so far, however, because it solely relies on data from its own ecosystem of platforms and first-party customer data provided by advertisers. Tools like Pixis and Smartly, which are integrated across social, search engines and shopping platforms including TikTok, YouTube, Google and Amazon, can use data from various sources to expand their targeting capabilities. Regardless of the optimisation tool used, the creative idea at the root of any campaign — and the creativity that underlies the different AI-made iterations — ultimately remains the most important piece of the puzzle. 'The original thought is the thing that will capture the human,' said Smartly's Marlow-Thomas. 'The artificial thought is the thing that will scale it and create lots of different iterations.'
