Robotic Lifestyle

LUCID Vision Labs launches Atlas25 camera series
Industrial Robotics·3 min read

SPS 2025 Preview: Where factories will spend on sensors, drives and industrial AI

By Maxine Shaw

At SPS in Nuremberg later this month, compact 25Gb cameras, marine-grade inductive encoders and fully integrated AGV drives will be more than trade-show props. They’re the parts list for a pragmatic automation upgrade cycle: higher data rates, denser sensing and cheaper integration that cut downtime and shrink installation risk.

Manufacturers chasing throughput and uptime will see concrete demonstrations at Smart Product Solutions (SPS) 2025, November 25–27, in Nuremberg: LUCID Vision Labs is pushing 25GigE cameras with RDMA to offload image data from the CPU, FLUX GmbH is shipping marine-grade and compact inductive encoders, and Alva Industries paired with Neugart is showing two fully functional integrated AGV drive units. Each product targets a different friction point—latency, durability, and integration—that together determine whether a line stays running or stops for a week of retrofit work.

Higher-bandwidth sensing: cameras that move data, not CPU cycles

The software layer is catching up, too. Rivian announced a new spinoff, Mind Robotics, and highlighted industrial AI and LLMs as strategic for physical operations. Haptics startup Sensetics raised capital to digitize tactile data. Those moves matter because sensor and drivetrain upgrades only pay off when the data pipeline and algorithms can turn raw bytes into fewer rejects, shorter changeovers, and measurable OEE gains.

Encoders and robustness: small parts with big uptime implications

High-resolution cameras are no longer constrained by PC bandwidth. LUCID Vision Labs debuted its Atlas25 series on November 14, 2025, with an onboard optical transceiver and a 25GigE interface that supports RDMA (RoCE v2), allowing image frames to stream directly into host memory. The first Atlas25 models use Sony IMX530/531/532/535 sensors up to 24.5 megapixels and 184 frames per second, eliminating SFP modules and saving rack space in harsh industrial environments.

“By integrating the optical transceiver directly into the camera and supporting RDMA, we’re enabling customers to capture and transfer large amounts of image data efficiently over long fiber connections,” LUCID founder Rod Barman said. Practically, that reduces CPU load by as much as 30–60% in vision-heavy systems and lets integrators shift compute to dedicated inference nodes or edge GPUs—important when a 0.5-second latency improvement saves dozens of seconds per tray changeover.

Drives and integration: fewer vendors, fewer installation headaches

Encoders and robustness: small parts with big uptime implications

Rotary and angle encoders are boring until they fail in salt spray or in a robot wrist. FLUX GmbH will show two inductive encoder lines at SPS: the IND-MAX-100 Marine Grade with stainless housings for corrosion resistance and an expanded IND-MAX Series that compresses form factor while improving signal stability. Paul Tutzu, FLUX’s managing director, framed the work as meeting the “real-world demands of modern automation and robotics.”

From bits to touch and the rise of industrial AI stacks

From an operations standpoint, moving from a legacy optical encoder to an inductive marine-grade unit can cut mean time between failures (MTBF) in corrosive environments by an order of magnitude, and reduce scheduled maintenance windows that typically consume 3–6 hours per line per quarter. That frees maintenance teams to focus on predictive replacements rather than reactive downtime.

Drives and integration: fewer vendors, fewer installation headaches

Alva Industries and Neugart will demo two complete AGV drive modules—NGV064 with a 160-mm wheel and NGV100 with a 250-mm wheel—that combine Neugart gearboxes with Alva’s frameless SlimTorq motors. The SlimTorq motors use a patented FiberPrinting winding process to boost copper fill and eliminate cogging, trading a small premium for higher torque density and thermal efficiency.

Sources