PR Newswire
Heavy‑industry operators have long struggled to reconcile the need for high‑precision automation with the reality of harsh field conditions. Conventional industrial PCs and server‑grade AI accelerators often require climate‑controlled enclosures, making them ill‑suited for open‑pit mines, demolition sites, or large‑scale construction projects. By marrying a ruggedized edge platform with a control system that can coordinate bipedal or dual‑arm robots in real time, ADLINK and Noble Machines aim to close that gap.
“The alliance addresses critical challenges regarding hardware durability and supply chain integration in extreme environments,” the joint release noted. “By seamlessly combining Noble Machines’ cutting‑edge with ADLINK’s Edge AI platforms, the partnership injects powerful momentum into the automation of heavy industry.”
If successful, the combined solution could enable operators to replace manual, high‑risk tasks—such as loading, material handling, and on‑site inspection—with autonomous robots that maintain productivity while reducing injury rates.
Technical foundation: DLAP Series + Jetson Thor
ADLINK’s DLAP (Deep Learning Accelerator Platform) series has historically targeted edge AI workloads in transportation, smart cities, and factory automation. The new iteration upgrades the compute heart to NVIDIA’s Jetson Thor, a system‑on‑module that delivers up to 150 TOPS of AI performance while maintaining a modest power envelope. This shift from the earlier Jetson Orin‑based designs reflects a deliberate emphasis on raw inference capability needed for whole‑body control loops that run at kilohertz frequencies.
- Extreme Environment Adaptability – The platform meets IEC 60068 standards for shock and vibration, and operates across a broad temperature range, ensuring reliability in dust‑filled, high‑vibration sites such as mines and demolition yards.
- High‑Bandwidth Sensing and Communication – Up to eight GMSL camera inputs (with Power‑over‑Cable), four Gigabit Ethernet ports, and optional 5G/Wi‑Fi expansion slots provide the data throughput required for multiple high‑resolution vision streams and low‑latency control signaling.
- Industrial‑Grade Power Management – Wide‑range voltage acceptance and advanced battery‑efficiency features allow the system to run longer on mobile power sources without sacrificing performance.
These hardware capabilities are intended to give Noble Machines the computational headroom to run its AI‑Driven Whole‑Body Control (WBC) stack, which coordinates locomotion, manipulation, and real‑time obstacle avoidance across both legs and arms.
The software side: Noble Machines’ Whole‑Body Control
Noble Machines, founded in 2024, has positioned its WBC stack as a universal layer that abstracts the complexities of multi‑limb coordination. The stack ingests sensor data from cameras, lidar, and force‑torque sensors, fusing them into a unified perception of the robot’s surroundings. It then feeds this information into a decision‑making module that selects motion primitives and generates joint‑level commands, all while maintaining balance and payload stability.
By integrating ADLINK’s rugged compute platform, Noble Machines can offload the most demanding inference tasks—such as object detection, semantic segmentation, and predictive dynamics—to dedicated AI accelerators, freeing the main control loop to maintain deterministic timing.
Executive perspectives
“Noble Machines’ leadership in the field of fully integrated AI‑Driven Whole‑Body Control and autonomy is undeniable,” said Ethan Chen, General Manager of ADLINK’s Edge Computing Platforms Business Unit. “Through this joint development agreement, ADLINK is rapidly introducing high‑performance computing hardware into the emerging general‑purpose robot market. Moving from our current DLAP platform support to a jointly developed dedicated platform featuring the NVIDIA Jetson Thor platform, we are committed to providing a long‑term, scalable computing architecture for the most challenging applications, accelerating Noble Machines’ entry into high‑value heavy industry markets.”
“Our alliance with ADLINK is not just a technical cooperation; it is a shared commitment to our industrial vision,” added Wei Ding, CEO of Under Control Robotics Inc. “ADLINK’s leadership in edge computing and their industrial‑grade design capabilities perfectly complement Noble Machines’ fully integrated AI‑Driven Whole‑Body Control and autonomy. This partnership enables us to deliver truly rugged, durable, and highly intelligent robotic solutions to our heavy industry clients, making the automation of dangerous tasks a reality.”
Market implications and competitive landscape
The heavy‑industry automation market, valued at over $70 billion globally, has historically been dominated by specialized, single‑function robots—such as robotic arms for welding or conveyor‑based material handlers. General‑purpose, mobile robots capable of handling varied payloads and navigating unstructured terrain remain a niche, largely because of the integration challenges between robust hardware and sophisticated control software.
By delivering a unified, rugged solution, the ADLINK‑Noble Machines alliance could pressure competitors like Boston Dynamics, Starship Technologies, and Kawasaki Robotics to accelerate their own hardware‑software co‑development efforts. Moreover, the inclusion of 5G and Wi‑Fi expansion hints at a future where these robots operate as part of broader edge‑cloud ecosystems, sharing sensor data for fleet‑wide optimization.
From a supply‑chain perspective, the partnership also mitigates a common pain point: the need to source separate compute modules, power‑management units, and communication adapters from multiple vendors. A single, integrated platform simplifies procurement, reduces integration risk, and shortens time‑to‑market—a crucial advantage for OEMs targeting projects with tight deployment schedules.
Real‑world rollout plans
The joint announcement indicated that a prototype built on the new platform will be showcased at GTC 2026, booth 141, in March. While detailed performance metrics were not disclosed, the presence at a high‑visibility NVIDIA event suggests that the partners intend to demonstrate live perception and manipulation tasks—likely moving heavy objects across a mock‑mine floor or assembling structural components on a construction mock‑site.
If the demonstration meets expectations, early adopters in sectors such as mining, construction, energy, and petrochemicals could begin pilot programs later in 2026, with broader commercial releases slated for 2027. The timing aligns with industry forecasts that predict a 30 % increase in robotic deployments for hazardous work by 2030.
Potential challenges
- Regulatory compliance – Deploying autonomous robots in public works or underground mines often requires adherence to strict safety standards and certification processes, which can extend rollout timelines.
- Integration with legacy systems – Heavy‑industry sites typically run on entrenched SCADA and MES platforms. Seamless data exchange between the new robots and existing infrastructure will be essential for ROI.
- Operator training – Transitioning from manual to robot‑assisted workflows demands upskilling of the workforce, a factor that can affect adoption rates.
Addressing these issues will likely involve additional collaborations with system integrators and industry bodies, an area where ADLINK’s extensive ecosystem could prove valuable.
Outlook
The ADLINK–Noble Machines partnership represents a concrete step toward bridging the gap between rugged edge AI hardware and the sophisticated control algorithms needed for truly autonomous, general‑purpose robots. By delivering a platform that can survive the harshest conditions while processing high‑resolution sensor data in real time, the duo positions itself to unlock new use cases in sectors that have been reluctant to adopt mobile robotics.
If the upcoming GTC showcase validates the performance claims, the collaboration could set a new benchmark for industrial robot design—one that emphasizes durability, compute density, and seamless integration over the more common approach of retrofitting consumer‑grade AI modules onto off‑the‑shelf chassis.
For enterprises grappling with labor shortages, safety concerns, and the need for flexible automation, the emergence of a rugged, Jetson Thor‑powered robot could shift the cost‑benefit calculus in favor of broader deployment. The next few months will reveal whether the technology lives up to its promise and how quickly the heavy‑industry market embraces it.
Get in touch with our Adtech experts
BusinessWire
