Deal focus: AI integration pays off for Evomotion

SenseTime might be China’s most valuable artificial intelligence (AI) technology developer, but its journey from private to public ownership was fraught with difficulty. A New York IPO was thwarted by a US government blacklisting in 2019 and more sanctions rocked a Hong Kong offering late last year. SenseTime persevered with a pared-back share sale and traded at a healthy premium to its IPO price.

Then, in late June, the share lock-up period expired, pre-IPO investors rushed to realise their gains, and the stock slumped by 50% in the space of 24 hours. It has yet to recover, and SenseTime posting a pandemic-influenced revenue contraction in the first half of the year hardly helps matters.

These trials are indicative of a subdued mood engulfing China’s once white-hot AI start-ups, with some investors now ending their participation in the space. Still, against this backdrop, China-based AI-of-things (AIoT) specialist Evomotion recently succeeded in closing a USD 20m Series B round. The lead investor was not disclosed, but COO Sheng Fu said it was an industrial rather than a financial player.

The capital-raising exercise – which began in the final quarter of 2021 – was buoyed by Evomotion’s solid financials. The seven-year-old company claims to be profitable and has worked hard to evolve from a pure-play software developer into a one-stop integrated solution, which is regarded as a more defensible strategic position.

“Software capabilities alone do not amount to a high enough barrier to entry. You need to integrate software with hardware to deliver a one-click solution or an off-the-shelf product,” said Simon Zhao, an executive director at Northern Light Venture Capital (NLVC). “Your product must be able to solve real problems – so your commercial capability, not just your technology, really matters.”

NLVC and Shenzhen-listed software developer ThunderSoft provided Evomotion’s angel round in 2015 and they returned for Series A and B rounds a year later, supported by Fortune Venture Capital. The company’s valuation has increased by more than tenfold since then.

The Series B represents Evomotion’s first external funding in six years, but the path to profitability has not been smooth. It started out developing computer vision algorithms and licensing them out to enterprise customers. However, it soon became clear that software wasn’t enough. The industry pain point was hardware – specifically chips – and the integration of software and hardware.

“It’s hard to acquire customers and make a profit just by offering software. System performance ultimately depends on hardware and how it is assembled, so software on its own couldn’t fill the gap between demand in each user scenario and the accuracy rate of the final product,” said Fu.

The integrated approach rests on close collaboration with brand owners, and from 2017, Evomotion started working with home appliance makers like Midea, Haier, and Joyoung. It helped washing machines identify clothes and choose the most appropriate washing mode, stoves detect smoke levels and adjust power levels, and ovens select cooking modes based on food types and monitor progress.

However, when appliance sales began to falter, brands focused on destocking and product development efforts were put on hold. Evomotion turned its attention to supermarkets. The problem it chose to address was self-checkout for fresh produce. Without scannable codes on packaging, customers would input codes manually, which was time-consuming and often resulted in confusion.

Evomotion used AI to automate the process and it has since achieved a near monopoly position in this field in China, working with general retailers such as Yonghui Superstores and RT-Mart, as well as many fresh-produce specialists.

Subsequent expansion has taken the company into livestock monitoring – chiefly serving lenders who use livestock as collateral for supply chain finance and agricultural insurers – and computer vision systems that enable robots to grasp non-standard items.

According to Fu, the company’s addressable market is not as large as smart city solutions, but there is less competition. The likes of Huawei Technologies and Hikvision are notable for their absence. In addition, US sanctions are less of a threat because the space is less conspicuous. Evomotion’s customers are mainly hardware integrators that buy modules with different features to form their own products.

• Tweet  
• Facebook  
• LinkedIn  
• Google plus  
• Save this article  
• Send to  

• Topics
• Greater China
• Early-stage
• Technology
• China
• Northern Light Venture Capital
• TMT
• artificial intelligence