High-purity PFA, a fluoropolymer essential for modern wet process equipment in the semiconductor industry, is gaining attention amid China's increasing drive to localize its semiconductor supply chain. This material is critical for manufacturing chemically resistant tubing, fittings, valves, and containers used in wafer cleaning, etching, and chemical-mechanical polishing processes.
The material's ability to resist contamination at an extreme purity level directly impacts the yield of functional chips on advanced technological nodes. The localization of PFA production within the country is occurring against the backdrop of expanding capacity by global memory chip manufacturers and accelerated investment in advanced logic processing.
Historically, the high-purity PFA market was controlled by several major global chemical companies with years of experience in producing ultra-pure fluoropolymers. The material must meet very stringent standards, including extremely low metal ion extraction, particle count control, and mechanical property stability in every batch. These requirements have proven technically challenging for new market entrants.
Several Chinese chemical companies have reached the batch verification stage. Their products undergo qualification testing by semiconductor equipment manufacturers and fab operators. The verification process typically takes several quarters and includes accelerated aging tests, chemical compatibility analyses, particle shedding measurements, and performance checks directly on the customer's equipment. Successful verification will allow Chinese equipment makers to procure this critical material domestically, thereby reducing reliance on foreign suppliers and shortening lead times, which can reach 12–24 months due to global equipment constraints.
The market potential is considerable, as every advanced semiconductor fab requires large volumes of high-purity PFA for chemical delivery systems, wet benches, and wastewater treatment infrastructure. As China builds numerous new wafer fabs under government-backed initiatives to achieve semiconductor self-sufficiency, the domestic PFA market volume is growing proportionally. Industry estimates suggest that demand for semiconductor-grade fluoropolymers in China could exceed several billion yuan annually over the next five years.
The technological complexity of this production cannot be underestimated. Producing high-purity PFA requires specialized polymerization processes, ultra-high purity raw materials, dedicated cross-contamination-free production lines, and sophisticated quality control systems capable of detecting impurities at parts-per-trillion levels. The verification stage is a crucial milestone, but achieving mass production with consistent quality at competitive prices may require additional investment in specialized manufacturing infrastructure. If local suppliers can successfully pass the verification and scaling phases, high-purity PFA could become another example in China's broader strategy of material localization in the semiconductor industry, which has already made significant progress in photoresists, etching gases, and CMP slurries.
In India, there is a problem of dependence on imported equipment, whether it is machines in factories, automation in hospitals or warehouses. Such equipment is often expensive to purchase, slow to maintain, and inaccessible to most enterprises.
iHUB Robotics, founded in Ernakulam, Kochi, in 2022, aims to bridge this gap by manufacturing humanoid robots in India from scratch: including the chassis, embedded systems, and artificial intelligence that controls them.
The company's journey took seven years. Three friends from Kerala—Athil Krishna, Akhil K Haridasan, and Sarat S—began researching humanoid robots in 2018, studying mechatronics in Coimbatore when the Indian ecosystem could barely support such work.
Despite this, they presented their ideas to the government, submitting defense proposals in 2019. In 2020, both the Border Security Force (BSF) and the PMO responded, with the BSF requesting a demonstration. However, the COVID-19 pandemic struck before they could create a working prototype, and the demonstration never took place. In 2022, they resumed work and officially launched iHUB, which is an abbreviation for Innovation Hub for Robotics, to turn research into a finished product.
The first product appeared in 2024—Tara Gen-1. This semi-humanoid service robot can speak over 100 languages, recognize faces and gestures, hold meaningful dialogues, and navigate a room autonomously using SLAM (Simultaneous Localisation and Mapping) technology.
It is released in various modifications: Tara Greet, Tara Learn, and Tara Care, which are adapted for the needs of hospitality, healthcare, education, banking, and public services. About 35 such units have already been deployed in three countries. This was followed by a heavier machine—Daksha, a general-purpose industrial humanoid capable of lifting loads up to 25 kg and operating at heights of eight to ten feet.
The foundation for both robots is what iHUB calls physical AI. Instead of the 'text-in to text-out' logic inherent in large language models, their robots operate on Vision-Language-Action models, which link what the machine sees with the language it understands and the movement it can perform.
The company creates its entire technology stack itself: from manufacturing and embedded systems to fine-tuning models and design, using an object in Kochi, Kalamassery. A corridor spanning about 24,000 square feet allows the robots to master tasks through repetition.
This vertical integration is a key advantage, as is the pricing policy, which allows systems to be used starting from approximately $10,000, significantly lower than the cost of imported humanoids. In January 2025, iHUB became the first Indian company selected for NVIDIA's Humanoid Robotics Program.
iHUB primarily sells its products to enterprises in sectors such as hospitality, healthcare, education, banking, manufacturing, and logistics, and also has a separate line for defense and security. The company has signed memorandums of understanding with EY Global and SAP Germany, as well as several Indian IT and industrial firms.
Tara Gen-1 has been exported to the UAE and Saudi Arabia. Currently, the company is financed by revenues from early projects and prototypes, rather than large external investments. In March 2025, it raised 4.3 crore rupees in a pre-seed funding round from American angel investors. These funds will go towards building the largest humanoid manufacturing plant in India, according to the company.
Future plans include scaling up production, securing research grants, and establishing an R&D center for industrial humanoids in Bengaluru, with plans to hire over 150 employees within two years. In addition to hardware, the founders intend to establish the iHub Training School to prepare students in robotics and AI.
The broader goal, although difficult to achieve, remains unchanged: India must produce humanoids independently, rather than importing them.
Despite significant hardware manufacturing in India, the country often relies on imported components. Computer modules and single-board computers, which serve as central elements for drones, industrial gateways, electric vehicles, and smart home devices, are mostly procured from abroad. This includes devices such as Raspberry Pi and Arduino boards, as well as foreign computing modules, putting Indian startups and manufacturers at risk of long lead times and uncontrollable overseas supply chains.
Edgehax, founded in Bengaluru in 2025, aims to eliminate this dependency by designing and manufacturing edge computing boards directly in India. The founders encountered this issue firsthand: Prabhu Stavarmath, CEO, spent over 15 years in IoT hardware, product development, and corporate sales, previously creating two hardware ventures—Bharat Pi and Refillbot—that served over 100,000 customers. Savitri Patil, COO, with experience in electronics and manufacturing, oversees operations, supply chain, and logistics. They concluded that the problem is structural: anyone producing a hardware product in India must start with an imported component responsible for its functionality.
Edgehax creates a unified board that integrates computing power, connectivity capabilities, and storage. This board is equipped with CPU, GPU, and NPU blocks and supports connections via 4G, 5G, Wi-Fi, and LoRa, featuring built-in storage. It runs on the company's proprietary operating system, DvarOS, specifically developed for industrial edge gateways. These boards are capable of executing convolutional neural networks and transformer models directly on the device, supporting remote deployment and software updates. The company asserts that the entire design and prototyping process is carried out in India.
The same hardware is used in diverse applications: industrial gateways, humanoid robots, autonomous vehicles, drones, NavIC-based tracking systems for India's defense needs, and consumer IoT devices. The customer receives a ready-to-use plugin with local sourcing and technical support, instead of a complex set of imported parts. While Raspberry Pi and Arduino are natural analogues for these boards, Edgehax positions its solutions as industrial, locally manufactured, and backed by long-term supply.
Edgehax has provided development kits to over 100,000 students and faculty across more than 30 universities and IITs, establishing a base of early adopters who have spread the product primarily through word-of-mouth and are now actively involved in developing, testing, and refining the kits. Commercially, the platform is utilized by over 150 startups, OEMs, and large enterprises, with over 5,000 edge gateway boards currently in production. Institutional support has been secured: Edgehax won the NXP Silicon Seeds Startup Program 2025 to create a low-cost computing module on NXP chips, and it also received the MeitY Bhashini award for developing equipment for the government's voice AI platform. Furthermore, the company is a beneficiary of the next-generation incubation program managed by Software Technology Parks of India, which provides mentorship and funding up to 25 lakh rupees to startups from smaller towns.
In August 2025, the company raised ₹1.39 crore in a seed round led by Inflection Point Ventures to fund production scaling, product development, and entry into international markets in Singapore, the US, and Europe in collaboration with NXP. This growth coincides with increasing global demand: global spending on edge hardware is projected to roughly double by 2030, reaching nearly $59 billion compared to $26 billion in 2025. The founders' goal is to become the primary design and manufacturing partner for OEMs requiring ready-to-use and customizable hardware, and to provide Edgehax boards and tools to 100 million developers in the Asia-Pacific region, the US, and Europe by 2029.
SpaceX is focusing on another test of the Starship, the company's largest rocket. This flight, number 13, is scheduled for July 16 and will feature the unprecedented transport of Starlink V3 satellites in an experimental mission.
According to the company, the launch is scheduled to take place at 6:45 PM in U.S. territory (corresponding to 7:45 PM in official Brazil time). The main purpose of this operation is to evaluate the vehicle's latest systems.
The departure will occur from the Starbase facility, located in Texas, near Boca Chica beach. This will be the second flight of the new V3 version of Starship. In addition to testing the rocket itself, the mission will serve as a demonstration for new components of the Starlink network.
The 20 Starlink V3 satellites to be carried by the spacecraft will not reach orbit. Instead, they will follow a suborbital trajectory and will be destroyed during the atmospheric reentry process, allowing SpaceX to analyze various phases of the flight.
SpaceX has a 90-minute window to conduct the launch. The official broadcast of the event will begin 30 minutes before the launch, made available through SpaceX channels on the X network and the official flight page.
The flight is expected to last a little over an hour and does not aim to take Starship into Earth orbit. The main focus will be monitoring the system's operation during crucial stages, such as module separation, engine activation, and vehicle return.
Key expected moments include: the launch and initial ascent of the rocket; the separation between the Super Heavy booster and the Starship vehicle; tests of the Raptor engines; the release of the 20 Starlink V3 satellites; and reentry followed by controlled splashdown in the ocean.
A point of great observation will be the performance of the Super Heavy booster. In the previous flight, this stage failed to return adequately for a controlled landing, resulting in its fall into the ocean. SpaceX reported that hardware and software adjustments have been implemented to resolve the issues detected in that mission.
The introduction of the new satellites represents one of the biggest innovations of this flight. The Starlink V3 models are larger and are part of the company's efforts to expand the capacity of the satellite internet network.
In a detailed description of the flight, SpaceX stated that 'for the first time, Starship will carry Starlink V3 satellites into space, which aim to significantly increase the network capacity and speeds for users.' Six of these satellites will be equipped with cameras to capture images of the heat shield and Starship during the mission.
The satellites will also attempt to establish communication with the Starlink constellation using high-capacity lasers before returning to the atmosphere. If the launch needs to be postponed due to weather conditions or technical problems, SpaceX already has a new planned date of July 17, maintaining the same time.
