Huawei Hubble Invests in Sight Robot's Angel Round; Team Led by 1998-Born PhD Develops Photonic Crystal Tactile Sensors

On October 13, Shenzhen Sight Tower Robotics Technology Co., Ltd. completed an angel round. Investors in this round include Huawei's Hubble Investment, a fund under Shenzhen High-Tech Investment, and Gaoxuan Investment, among others, with Duowei Capital serving as exclusive financial advisor. The source did not disclose the specific amount, valuation, or equity percentage of this round.

What made industrial capital take a stake was not a smarter algorithm team, but a route choice originating in materials science. Sight Tower founder and CEO Zhang Wei was born in 1998 and was jointly trained during his doctoral studies by Southern University of Science and Technology and City University of Hong Kong, with research spanning optical materials and robotics. The company has a team of about 50 people, averaging about 23 years old, with core members from the Southern University of Science and Technology Intelligent Liquid Crystal Optics Laboratory, covering fields including materials, device circuits, and algorithms.

The problem awareness behind this route comes from a specific gap: vision can tell a robot where an object is, but it is hard for vision to accurately tell the robot, after contact occurs, how much force was used or whether slipping occurred. Technical routes such as piezoresistive, capacitive, magnetic tactile, and visuotactile have developed for years, each with applicable scenarios, but achieving thinness, high density, and low cost at the same time within limited spaces such as robot fingers and palms still involves many engineering difficulties. Sight Tower chose to look for the answer in the material itself.

Sight Tower's core technology is a class of flexible photonic crystal materials, represented by cholesteric liquid crystal elastomer (CLCE). The color of this material is not dyed by pigments, but produced by the periodic helical structure formed by internal liquid crystal molecules. When the material is subjected to pressure and deforms, the period of the internal helical structure changes accordingly, which in turn causes changes in the reflection spectrum and color.

The source gives an intuitive quantitative example: after photonic crystal structures of six to seven hundred nanometers are stacked ten layers, a one-micron change can make the wavelength go from 600 nanometers to 500 nanometers, and the color go from red to green. This is equivalent to using the material's optical response to express originally tiny deformation changes as obvious color differences. This example is a source description and no test conditions were provided.

For Sight Tower, the real technical work is not discovering that photonic crystals can change color, but how to make this material into a tactile sensor that can be stably read, repeatedly manufactured, and ultimately integrated onto a robot. The source says Sight Tower's mass-produced sensors underwent 1 million tests, with error rates all below 5%. This figure is the company's internal testing basis, and the test conditions and comparison benchmarks were not disclosed.

Sight Tower's next-generation electronic skin solution combines mechanochromic flexible photonic crystals, a transparent intermediate layer, a flexible TFT photosensitive array, and a backlight layer. Among these, the photonic crystals are responsible for sensing contact and deformation, and the flexible photosensitive array is responsible for reading optical signals. Unlike traditional visuotactile approaches that need to reserve imaging space for a camera, this design integrates the optical reading device into the thin-film structure, offering the possibility of further compressing sensor thickness.

The source says Sight Tower is combining flexible photonic crystals with photosensitive devices to develop a flexible electronic skin solution less than 1 mm thick with a sensing density exceeding 10,000 points/cm². These two figures correspond to two key questions: first, whether it can fit, since robot fingers and palms have limited space and a sensor that is too thick will affect the body structure design; second, whether it can sense finely enough, since contact during grasping is often distributed across multiple areas, and higher spatial resolution helps record contact details.

What needs to be distinguished is that being less than 1 mm thick and having a sensing density exceeding 10,000 points/cm² are Sight Tower's solution descriptions; the source did not clearly distinguish whether these are solution goals or already delivered products, nor did it provide third-party testing or mass-production validation data. Flexibility further expands the application range: the human hand is not flat, and fingers, knuckles, and palms all have complex curvatures, and a flexible thin film has the opportunity to spread along these surfaces.

Being able to bend does not mean that a complete electronic skin already has mature large-area deployment capability. The source clearly points out that the photosensitive array, packaging structure, device consistency, and long-term durability all need to continue to be resolved in subsequent engineering. This is also an important direction for the next stage of Sight Tower's electronic skin: letting touch expand from the fingertip to a larger area of the hand surface, and further exploring tactile coverage of the full palm and even other parts of the robot body. The source did not give a timetable for achieving this coverage.

From laboratory material to deliverable product, there are still many engineering problems in between: how to ensure material consistency, how to package the sensor, how to reduce manufacturing costs, and whether new customer requirements for size and structure can be adjusted quickly. Sight Tower has about 800 square meters of R&D and production space in Shenzhen, equipped with a cleanroom, 3D printing, CNC, laser processing, circuit mounting, and other equipment, as well as a test platform built by the team itself. This model of tightly linking R&D and manufacturing allows engineers to complete design, processing, and testing at the same site, shortening the validation cycle.

Less than a year after founding, Sight Tower has gradually established an R&D and delivery system for tactile sensors and data-collection equipment, with customers covering university laboratories at home and abroad, leading model companies, and dexterous hand manufacturers. The source did not disclose the specific customer list or procurement scale.

During commercialization, Sight Tower found that different customers have clearly different requirements for touch: some focus on response speed, some pursue higher force measurement accuracy, some need to sense texture, and some are first limited by installation space. This means that even if photonic crystals have unique advantages in thinness and high-density sensing, it does not mean they can replace all tactile technology routes.

Therefore, while Sight Tower continues to advance R&D on flexible photonic crystal electronic skin, it is also laying out different tactile approaches such as visuotactile, piezoresistive, and capacitive around customers' actual needs and providing corresponding sensor products. This parallel multiple-route approach is both a commercial compromise and an indication that photonic crystal electronic skin is still currently a core R&D direction rather than the only delivery form.

In data collection, Sight Tower has developed devices such as a visuotactile data-collection gripper, an EGO data-collection system, and a data-collection backpack, supplementing information such as contact states and force changes for robot model training. Beyond hardware, the company is also extending toward robot data and operation control.

Sight Tower is developing its own tactile control model, an "external tactile cerebellum." Its design idea is to coordinate with upper-layer robot models such as VLA: the upper-layer model is responsible for understanding tasks and planning actions, while the tactile model makes local adjustments to force, position, and posture based on contact feedback.

The source says that in internal tests on some typical tasks such as erasing a blackboard and inserting into a hole, after introducing tactile cerebellum feedback, robot task success rates improved by up to nearly 100%. This figure is the company's internal testing basis, and the test scale, task definitions, and statistical basis were not disclosed; it should be regarded as a source claim rather than an independently verified conclusion.

The joint development of sensors and models allows Sight Tower to start from actual operation needs and, in turn, optimize hardware design and data collection methods. However, flexible photonic crystal electronic skin is still Sight Tower's most core product direction at this stage. Compared with a single fingertip sensor, electronic skin has the opportunity to cover a larger contact area, and it also places higher demands on material preparation, device integration, and large-scale manufacturing.