Artificial intelligence (AI) is an evolving technology that is still growing, but it is undoubtedly getting better. 

For all we know, factories might not even need lights 20 years later, because most of them will be operated by AI. We see a lot of R&D happening within the AI framework, giving good results. Besides, we see newer frameworks coming in. 

AI-based visual inspection, too, has been growing by leaps and bounds, reshaping automotive inspection. It helps examine detailed defects in vehicles, providing automotive OEMs the opportunity for accuracy and cost-savings. 

One such company focusing on AI-based visual inspection is Lincode Labs, whose AI-backed visual inspection solution, Lincode Visual Inspection System (LIVIS), is its current focus. The company was started with complete research and understanding of the top challenges that manufacturers globally face. After interviewing close to 100 customers, 86 percent of them said that quality inspection happens to be their biggest challenge.

“We were intrigued by this and went to various quality inspection people and identified the technology they were using,” says Rajesh Iyengar, Founder and CEO, Lincode Labs, and goes on, “We went on to find out that the technology hasn’t changed for two decades and there were a lot of false calls in it. That’s when Lincode stepped in and built a product around specific challenges focused on the automotive industry.”

Automotive OEMs, too, look at specifically reducing these false calls and improving productivity, which Lincode has helped solve through its AI-backed visual inspection solution. “The industry standards were 150 to 200 false calls per million inspections. So, in our case, we are doing it in zero to four parts per million,” Iyengar cites.  

Iyengar further reveals that due to this, 80 percent of their customers are repeat orders. “That’s because they are completely happy with the inspection process and the way the inspection is automated,” he mentions.  

LIVIS
Traditional vision systems cannot catch up with AI, as Iyengar says. He avers, “LIVIS stands out because we have built it as a platform. The scalability becomes easier if you’re going to deploy it across multiple factories and locations. But also, the foremost important thing is that it is completely made as a product. Thus, AI is commoditised. With the LIVIS platform, we can bring the cost and time to deployment down.” 

Lincode’s role in the automotive industry
What’s interesting is that even if Lincode caters to the manufacturing industry as a whole, it first addressed the automotive industry. The company researched the market size of various manufacturing companies and the automotive industry took the top spot, with close to USD 542 billion of global value.  

“We started with the automotive industry but pivoted later,” Iyengar tells us and continues, “So, instead of looking at just the automotive or any other industry, we turned our attention to steel, metal, plastic, glass etc. We basically went to the surface and saw how steel and metal are produced today. Whether it’s a CNC machining or forging or casting process, these are major processes used for any industry across the globe involving steel and metal. We understood that steel and metal are dealt with in the same way globally. Therefore, it made sense to go to the surface and into these kinds of defects specifically, and then generalise that and start building a model towards it. This, plus making AI as a product, has made deployment easier across the globe.” 

R&D centre in Bengaluru
Lincode recently opened a new R&D centre in Bengaluru, which also has a significant role to play in deploying the company’s solution across the world.

Stressing on the fact that evolving models are important in AI, Iyengar states, “It’s a continuous process; it’s not that you just build a model and you’re set. We have a big roadmap in the product development, and the Bengaluru R&D centre is going to play a major role in that. We are going to conduct deep research with various data collected across the globe and do various testing with that.” 

Staying ahead
What’s more, Lincode recently closed a funding round in December last year. Catering to a constantly evolving industry like the automotive, Lincode, too, strives to make sure that its visual inspection solution stays ahead and is put to use. “There are about more than 600 parts in a car and each part is segregated – like the structure, wiring, engine components etc.,” Iyengar shares and continues, “These segregations are made so that we can target the sector of the product. For example, when it comes to engine blocks, there is a specific model with a huge data set around engine blocks. This is how we stay ahead of competition.” Iyengar also adds that their trials with various use cases made them understand that inspection alone is not important but also the way the inspection is done. 

Essential skill sets for AI vision systems
Leveraging AI-based visual inspection solutions in the automotive industry is bound to increase productivity, and the cost of labour will also come down because of automation. “Today, most manufacturers use secondary inspection, which can be cut off straight away. This will improve their productivity and also reduce the risk of delays,” Iyengar enlightens. 

Moreover, AI vision systems come with their share of essential skill sets to bring out the best in the automotive industry. Iyengar states that, in general, skilling is required for the factory people. “This could be at various levels,” he puts across and adds, “It could be for the operators, the IT administrator or even the software development team. Hence, deep training is required, which can be somewhat cumbersome because it could be a bit challenging for the operator. So, an IT person might be needed in order to help the operator every time there is a downtime.” 

Covid-19 and AI-backed visual inspection
Such training or skills could certainly come in handy, because Iyengar claims that the need for AI-backed visual inspection solutions in the automotive industry has increased since the Covid-19 pandemic. “Unplanned shutdowns happened during Covid, because of which employees could not report and manufacturing could not continue properly,” he responds and adds, “Hence, a lot of investments are happening because of this. In fact, even now, a lot of employees are still not reporting and the labour problem has become global. It has become tough to get skilled workers. This has led to the adoption of autonomous manufacturing for automotives, where AI is going to play a big role.”

Meeting industry requirements
For an industry that is an economic force globally, AI-based visual inspection is certainly meeting the high-quality requirements of the customers of the automotive sector. Plus, we already see companies like Volvo using the technology. Safety surpasses any requirement, and this requirement can be fulfilled if quality is top-notch. And quality will be at its best if automotive manufacturers can perform production quality inspections in the most efficient way. (MT)

Comments (0)

ADD COMMENT

Link Copied!

Artificial intelligence software developer Helm.ai has launched two foundation models, GenSim-3 and VidGen-3, establishing a native Full HD (1920x1080) resolution standard for generative simulation across a 6-camera, 360-degree surround-view suite.

The architecture delivers 5x the pixel density of industry benchmarks to assist automotive developers facing the limitation where physical collection of edge cases becomes logistically restrictive.

Traditional generative world models typically cap resolution at roughly 0.4 megapixels per camera. Helm.ai’s platform outputs a native 2 megapixels per camera, yielding a synchronised 12-megapixel synthetic canvas per timestep. This specification matches the hardware parameters of production-grade vehicle cameras to reduce the domain gap for SAE Level 2 through Level 4 autonomous vehicle development.

The platform functions as a virtual sensor twin by mathematically replicating physical constraints and hardware anomalies, including lens flares, sensor banding patterns, and exposure blinding. To accommodate different neural network training routines, the pipeline can be configured to a high-speed validation mode using a three-camera setup at 30 frames per second, or a spatial context mode generating a six-camera surround view at 5 frames per second.

Data generation is split into two operational pipelines. GenSim-3 focuses on data augmentation by modifying environmental parameters such as weather, lighting, and object surfaces across real-world video segments at native 2MP resolution. VidGen-3 focuses on data creation, synthesising driving sequences from scratch by simulating environments, agent behaviours, and traffic logic without baseline video to patch geographic data gaps.

Helm.ai achieved the 2MP standard using a cluster of a few hundred GPUs rather than the thousands typically required for sub-HD video generation. This framework reduces the GPU infrastructure footprint for vehicle manufacturers and provides a method for compressing autonomous driving software onto mass-market on-vehicle compute chips.

Vladislav Voroninski, CEO and Founder, Helm.ai, said, "We are moving the industry from standard 'AI video' to authentic, hardware-accurate sensor emulation. By leading with a Full HD (2MP) standard and a 12-megapixel total aggregate capability per timestep, we have solved the resolution bottleneck that has historically limited the utility of generative AI in safety-critical systems. By optimising our compute architecture, we are giving our partners a high-performance platform to validate their autonomous stacks using synthetic data that perfectly matches the fidelity of their actual production sensors."

Comments (0)

ADD COMMENT

Link Copied!

Global automotive technology supplier Marelli has marked the 30th anniversary of its flagship electronics manufacturing plant in Guangzhou. Established in 1996 as Marelli’s inaugural manufacturing investment in China, the facility has transformed from a baseline assembly outpost into a major smart manufacturing and hardware-software validation centre.

Over the past three decades, the facility has expanded from a single operational production line with approximately 100 technicians into a 30,000-square-meter automotive electronics campus.

Today, the facility employs nearly 1,000 people and runs 66 active production lines, manufacturing components for both localised Chinese vehicle programs and global vehicle architectures.

The campus houses an adjacent, fully integrated Engineering Center that holds more than 100 registered patents. The manufacturing framework integrates high-precision assembly lines, automated optical bonding modules and site-wide rooftop solar arrays designed to manage factory energy overheads and lower operational carbon density.

The Guangzhou plant functions as a strategic industrialisation hub focused on low-cost, scalable architectures suited for the industry transition toward connected, software-defined vehicles (SDVs). The facility specialises in several high-growth hardware and display segments like advanced display solutions based on Mini-LED and MicroLED technologies. Additional key platforms include electronic control units (ECUs) for body and seat systems, zone control units, as well as digital cockpits, digital instrument clusters, and 5G telematics systems.

Ravi Tallapragada, President of Marelli’s Electronics business unit, said, “Our Guangzhou plant is a cornerstone of Marelli’s Electronics business in China and a powerful example of how innovation and advanced manufacturing can drive sustainable growth. Over the past 30 years, the team has continuously evolved its capabilities, developing advanced technologies and scalable platforms that address the rapid transformation of the automotive industry, building on long-standing collaboration with customers and partners. I’m proud of our team in Guangzhou and confident that the plant will continue to play a key role in shaping Marelli’s future globally.”

Comments (0)

ADD COMMENT

Link Copied!

Bosch MC Battery Service Innovations, a 50:50 joint venture established by Bosch and Mitsubishi Corporation, has secured its first commercial customer for its ‘Battery-as-a-Service’ (BaaS) solution. The platform has officially gone live with the opening of an automated energy service hub in Chizhou, Anhui Province.

The new logistical hub is owned and operated by Shanghai Lingzhou Technology Co. The site acts as a high-volume transit point configured specifically for heavy-duty electric trucks, allowing operators to either swap out spent battery packs or utilise fast-charging bays within a few minutes. The Chizhou station currently processes more than 100 commercial electric trucks per day.

The commercial roll-out aligns with rapid fleet electrification trends in China, where New Energy Vehicles (NEVs) accounted for nearly 30 percent of all heavy-duty truck sales in 2025. Internal market projections from Bosch indicate that over 50 percent of new truck registrations in the country will be fully electric by 2030.

The joint venture's business model separates the acquisition cost of the vehicle chassis from the battery chemistry, mitigating a core hurdle for corporate logistics fleets: accounting for unpredictable battery degradation and its subsequent impact on total cost of ownership (TCO) and residual vehicle asset valuation.

The operational framework of the joint venture utilises a collaborative technical and commercial division between both partners. As per the understanding, Bosch will provide the core software stack for the platform. Real-time operating metrics – including local ambient temperature profiles, instantaneous current loads and historical charging frequencies – are beamed continuously to cloud servers. The algorithms evaluate the exact state of health (SoH) of individual packs, run predictive wear modelling to catch cell stress anomalies and dynamically manage fast-charging protocols to minimise thermal strain.

On the other hand, Mitsubishi manages localised market deployment, regulatory anchoring and downstream financial underwriting. The data harvested through the tracking platform is being funnelled into integrated aftermarket networks to build commercial products, including predictive hardware maintenance contracts, connected usage-based fleet insurance packages and secondary-life battery storage applications.

Thomas Pauer, President of the Bosch Power Solutions division, said, "With this service, Bosch and Mitsubishi Corporation can create real added value for fleets. Although the state of health can decline due to ageing and many charging cycles, our solution allows fleet operators to keep an eye on the battery condition of their vehicles – a decisive criterion for the everyday suitability and total cost of ownership of a fleet."

Qian Yang, General Manager of the joint venture’s local subsidiary in China, said, "Our service hits a local nerve: We support battery-electric vehicles in the fleet business. This holistic approach accelerates the electrification of fleets and optimises the entire battery lifecycle. The combined expertise of Mitsubishi and Bosch is a perfect match for our customers."

Comments (0)

ADD COMMENT

Link Copied!

L&T Electronic Products & Systems (LTEPS) has formed a strategic partnership with Israeli electric propulsion technology firm EVR Motors to co-develop, manufacture and distribute next-generation electric vehicle traction motors for the Indian market.

The collaboration will focus on industrialising traction motors optimised for regional conditions. The designs are engineered for high operational efficiency, compact physical packaging and a reduced reliance on rare-earth materials. The product line will cater to multiple transport segments, ranging from electric two-wheelers and three-wheelers to passenger cars and heavy commercial vehicles.

The traction motors will be manufactured at the LTEPS production facility in Coimbatore, Tamil Nadu. When paired with LTEPS's indigenously designed Motor Control Units (MCUs), the combined hardware will provide automotive original equipment manufacturers (OEMs) with a complete, integrated EV powertrain and drivetrain solution.

LTEPS is a business unit of Larsen & Toubro that specialises in the design, engineering and manufacturing of high-reliability electronic systems across the aerospace, defence, industrial and energy sectors.

EVR Motors specialises in proprietary electric motor topologies, utilising a patented Trapezoidal Stator radial flux architecture designed to reduce weight, size and material volume while maximizing power density.

Prashant Jain, Head of L&T Electronic Products & Systems, said, “This partnership reflects our commitment to developing indigenous, high‑performance solutions that support India’s clean mobility ambitions. By combining advanced motor innovation with indigenous motor control unit, we bridge the gap between cutting‑edge technology and real‑world deployment across India’s EV landscape.”

Opher Doron, CEO, EVR Motors, stated, “Our collaboration with LTEPS enables us to scale innovation responsibly – delivering traction motor solutions that are not only technologically superior, but also manufacturable, reliable and tailored for Indian mobility needs.”

Sajal Kishore, Managing Director, EVR India, added, “India’s electric mobility transformation requires system-level powertrain integration and deep localisation. The collaboration between EVR Motors and L&T will enable next-generation electric powertrain solutions across mobility segments.”

Comments (0)

ADD COMMENT

Link Copied!