SHANGHAI, July 18, 2026 /PRNewswire/ — The 2026 World Artificial Intelligence Conference (WAIC) opened in Shanghai on July 17. Shanghai Blacklake Technologies Co., Ltd. (“Black Lake”), an industrial AI company, is showcasing a portfolio of industrial AI agents at the conference. The company has also been named to the Top 30 shortlist for the 2026 WAIC Super AI Leader (SAIL) Award and selected as a Trusted Partner under the Global Call for Trusted Partners for Industrial AI in the Global South.

The accreditations highlight Black Lake’s latest progress in bringing AI into critical manufacturing decision-making workflows and deploying industrial AI capabilities on the shop floor around the world.

This year’s conference attracted over 1,100 exhibiting companies and showcased more than 3,000 exhibits, setting a new record for exhibition scale. The conference delivered a clear signal: as artificial intelligence becomes a common priority across global industries, attention is moving beyond model capabilities toward practical applications in real-world operating environments.

Manufacturing provides a particularly demanding test for this transition. Factory operations are governed by multiple constraints, including process specifications, equipment capabilities, material availability, production capacity, delivery schedules and quality requirements. Therefore, AI has to do so much more than simply comprehend information input. It must make reliable judgments within clearly defined business rules and operational constraints.

Black Lake has focused on industrial digitalization and industrial AI for years, developing and deploying AI applications in a range of factory environments.

At WAIC 2026, the company is presenting industrial AI agents covering order splitting and process planning, quotation and pricing, procurement, production scheduling, quality inspection, and order tracking. These applications are designed to move AI beyond an auxiliary role and into critical manufacturing decision-making workflows.

Traditional industrial software is primarily responsible for data recording, digital workflows, and worker coordination. However, critical decisions such as how to split an order, determine pricing, schedule production, and assess quality risks still depend heavily on the experience of engineers and frontline workers.

Industrial AI agents are intended to convert fragmented industrial knowledge and production experience into decision-making capabilities that can be invoked, reused and continuously refined by software systems.

Order decomposition and process planning are representative examples. After receiving an engineering drawing, a factory typically relies on experienced engineers to identify components, materials and dimensions, define the required manufacturing processes and technical specifications, and establish a basis for subsequent quotation and quality inspection.

The process is highly dependent on individual expertise and represents one of the first critical decision points after an order is received.

Black Lake Technologies’ CAD-to-Process Agent can understand product drawings and, taking into account the factory’s equipment capabilities, process requirements, and production practices, rapidly generate process steps along with the corresponding technical requirements. Drawing analysis that once took hours can now be completed in approximately one minute, achieving an accuracy rate of over 95% in real deployment and providing engineers with stable, efficient decision support. Currently, the industrial agents developed by the company cover core processes including design, scheduling, production, and quality inspection, and have entered the stage of large-scale deployment.

Founded in 2016, Black Lake serves nearly 40,000 factories worldwide. Its customers span more than 30 industries, including food and beverage, automotive components and equipment manufacturing.

By working across factory order management, production and fulfillment workflows, Black Lake has accumulated the technical capabilities and industry knowledge required to support decision-making in complex industrial environments.

In April 2026, Black Lake completed a Series D funding round of nearly RMB 1 billion. The company said the proceeds would primarily be used to accelerate the deployment of its industrial AI products and support its international expansion.

AI-related products are becoming a new source of growth for the company. In a recent interview, Black Lake founder and CEO Zhou Yuxiang said that the company had recorded significant growth in AI-related revenue since the beginning of 2026. He also said that manufacturing customers were taking less time to make purchasing decisions for industrial AI agents.

Zhou expects AI adoption among Chinese factories to increase substantially over the next three to four years.

Unlike consumer-facing AI, which is primarily associated with content generation and personal productivity, industrial AI agents can directly affect production costs, capacity utilization, delivery performance, and product quality. Their commercial value therefore depends largely on whether they can perform specific tasks reliably in complex production environments.

During WAIC 2026, Black Lake was named to the Top 30 shortlist for the 2026 Super AI Leader (SAIL) Award. The SAIL Award is one of WAIC’s major awards and recognizes achievements in technological breakthroughs, application innovation, and industrial value.

Black Lake was also selected as a Trusted Partner under UNIDO’s Global Call for Trusted Partners for Industrial AI in the Global South.

The Global Call was launched under the guidance of the United Nations Industrial Development Organization (UNIDO), in partnership with the Shanghai Artificial Intelligence Research Institute, and in connection with the work of UNIDO AIM Global and its Shanghai-based Centre of Excellence.

The initiative aims to build a curated pool of leading partners to co-develop scalable industrial AI solutions and public goods for the Global South.

For Black Lake, the two accreditations underscore the growing importance of reliability, explainability, and scalability in the evaluation of industrial AI, in addition to the capabilities of AI models.

Global expansion will be a major priority in the company’s next phase of development. Black Lake is currently focusing on Southeast Asia, Latin America and Eastern Europe, adapting its industrial AI agents to the industrial structures, production processes and management requirements of different markets.

Although manufacturing operations vary across countries and regions, manufacturers share similar concerns about efficiency, quality, delivery reliability and production flexibility.

Black Lake is transforming industrial AI capabilities that have been validated in complex factory environments into configurable and deployable products. Through these products, the company aims to work with manufacturers worldwide to explore more efficient, flexible and intelligent approaches to production.

SHANGHAI, July 17, 2026 /PRNewswire/ — Joyson Electronics (600699.SH/0699.HK) announced at the 2026 World Artificial Intelligence Conference (WAIC) a suite of robotic component solutions, including a dexterous robotic hand, solid-liquid hybrid battery, third-generation AI head assembly, electronic skin, and an embodied AI brain – alongside its latest achievements in industrial settings training and application. In addition, the company’s robot controller products are already in volume production and being delivered to leading robotics firms.

Dexterous Robotic Hand Integrates Multiple Industry-Exclusive Technologies; AI Head Assembly Ready for Rapid Mass Production

The dexterous robotic hand is often regarded as the “crown jewel” of robotics – owing to its high level of integration across a broad range of frontier disciplines, among them bionics, flexible sensing, MEMS, and advanced materials – and its significant commercial value.

At WAIC, Joyson Electronics introduced its first fully in-house developed “TeleHand” series of dexterous hand solutions. The TeleHand Professional Edition features an industry-exclusive “in-palm integration + hybrid actuation” architecture, directly addressing key challenges such as standalone integrity, tactile sensing, compliant manipulation, and fine motion control.

With 20 degrees of freedom, the TeleHand integrates three actuation modes – direct drive, tendon-driven, and linkage – within the palm. This design not only combines the precision of direct drive with the compliance of tendon-driven mechanisms, but also delivers higher transmission efficiency and lighter weight, enabling easy adaptation to various robotic platforms.

The TeleHand is equipped with Joyson Electronics’ in-house developed actuators and force-tactile sensing technologies, including:

• Ultra-compact, high-torque-density miniature frameless actuators, which reduce volume by nearly 50% and weight by approximately 30% compared to conventional models, while delivering 2–3 times higher torque density than industry-standard hollow-cup motors of the same diameter.
• In-house developed force and tactile sensing technology (electronic skin), featuring industry-exclusive natively decoupled three-dimensional force sensing, achieving resolution beyond human tactile limits, with high sensitivity, proximity detection, ultra-thin form factor, and flexibility – suitable for diverse dexterous hand and embodied intelligence applications.

The TeleHand PHINO platform’s native unified multimodal fusion architecture minimizes information loss and offers strong generalization capabilities, enabling the TeleHand to perform precision industrial operations while seamlessly supporting service-oriented interactive scenarios. In addition to the Professional Edition, Joyson Electronics also launched a cost-effective Basic Edition, which offers industrial-grade reliability and real-world deployment advantages through in-house factory batch deployment.

Meanwhile, Joyson Electronics unveiled its third-generation AI head assembly, which integrates perception, motion, and system-level capabilities to deliver more natural head movements and emotional expression. Designed with a production-ready mechatronic architecture, it enables rapid support from concept design and prototype validation to mass production. Its modular and platform-based design further allows for agile product customization and iteration to meet diverse customer requirements.

Joyson Electronics Debuts Embodied AI Brain; Controllers Already Shipping to Leading Customers

In the robotics “brain” domain (cerebrum and cerebellum), Joyson Electronics’ automotive-grade edge-side physical AI platform – its robot controller products – has already achieved commercial deployment and is now in volume production for leading robotics customers.

Furthermore, Joyson Electronics unveiled its embodied AI brain solution (EAOS + EAPC) – a unified, software-hardware-integrated platform designed for cross-form-factor and cross-scenario adaptability. The solution aims to make robots “easier to use, truly productive, and capable of autonomous evolution.”

The Embodied AI PC (EAPC) adopts an external form-factor design, built on a fused cerebrum-cerebellum controller architecture, with computing power ranging from 40 TOPS to 2070 TFLOPS, meeting diverse requirements from entry-level to flagship embodied AI systems. The product features a modular, integrated design with a compact footprint and superior thermal efficiency, enabling cross-platform and cross-environment adaptability. Leveraging Joyson Electronics’ automotive supply chain and manufacturing capabilities, the solution also offers significant cost competitiveness.

On the software side, the Embodied AI Operating System (EAOS) comprises three core subsystems:

• World Model – responsible for “understanding”, encoding multimodal signals into unified state representations and using dynamic predictors to simulate and preview scenarios within the system.
• Agentic OS – responsible for “action”, formulating high-level strategies, decomposing complex tasks, dynamically orchestrating sub-agents, invoking skill libraries and tools, and translating decisions into precise motions across dexterous hands, robotic arms, and mobile chassis.
• Memory System – responsible for “evolution,” managing working memory for real-time context, episodic memory for past experiences, and skill memory for accumulated learned capabilities.

The EAOS enables robots to execute long-horizon, complex tasks and achieve autonomous evolution – translating into tangible productivity gains. To date, Joyson Electronics’ embodied AI brain has been deployed in real-world settings, including select industrial scenarios and automated charging.

Solid-Liquid Hybrid Battery: The Optimal Power Solution for Embodied Intelligence

Conventional energy solutions for embodied intelligence face multiple challenges – limited endurance, large footprint, long recharging times, and insufficient power capacity to support instantaneous high-current discharge. Battery safety also remains a critical factor for widespread adoption. The industry requires a fundamental breakthrough that simultaneously balances energy density, power density, and safety.

Solid-liquid hybrid batteries (semi-solid-state batteries) offer the optimal power solution for embodied intelligence and represent the only technological pathway capable of addressing all the above energy challenges at the current stage. At WAIC, Joyson Electronics introduced its “Crystal Energy” multi-form solid-liquid hybrid battery solution, delivering high performance and reliability:

• Energy density significantly increased to 380 Wh/kg
• Overall endurance improved by approximately 60%
• Cycle life exceeding 2,000 cycles
• Wide operating temperature range from -20°C to 60°C
• Supports both wired and wireless charging, reaching 80% capacity in just 30 minutes

Complementing this is the Crystal Energy Ultra-Control BMS, which operates across a wide temperature range of -40°C to 105°C, featuring real-time cell monitoring, automotive-grade safety protection, and full-lifecycle health management – comprehensively enhancing the safety, durability, and energy efficiency of robotic power systems. Additionally, Joyson Electronics unveiled its first gallium nitride (GaN) motor driver, achieving conversion efficiency exceeding 95% while reducing size by 40% – positioning it at the forefront of the industry.

With robotics standing on the cusp of large-scale commercialization, Joyson Electronics is advancing its “self-development + investment” dual-drive strategy, expanding its presence in embodied intelligence, and accelerating breakthroughs in key technologies. Looking ahead, Joyson Electronics will leverage its global R&D, manufacturing capabilities, and industrial settings to drive the reliable, scalable, and cost-effective commercialization of robotic core components, helping accelerate the industry’s transition to mass adoption.