From the assembly line to the black light lab, the automation revolution upstream of the life sciences

People who have done experiments probably can't escape these: writing endless experimental reports, complex experimental instruments, cumbersome experimental data processing, and difficult to repeat experimental results... It has been a bitter "experiment" for too long.

In recent years, these pain points are gradually being broken. In 2019, the University of Liverpool research team invented an AI robot chemist, and in early 2022, it independently discovered a highly active catalyst for the first time, and the laboratory entered a new level of "intelligence". Pull the perspective back home. At the same time, the new coronavirus epidemic has made large-scale nucleic acid testing a necessity for epidemic prevention, and the batch emergence of innovative biomedical companies has also sharply increased the demand for relevant research and development experiments. For this reason, the "unpopular" industry of laboratory automation is heating up rapidly.

The strength of the capital is the most intuitive reflection of the industry heat. Summarizing the institutions that have invested in the laboratory automation industry in the past two years, Sequoia China, Hillhouse Venture Capital, IDG, Qiming Venture Capital, Northern Light Venture Capital, Sinovation Works, Matrix China, Bytejump, Shunwei Capital and other star VCS and war investment have been listed, both TMT funds, and pharmaceutical industry funds. There are also CXO (Wuxi Apptec), biochemical analyzer (Perkinelmer), automation (Bosch) and other fields of industry leaders, everyone smelled the opportunity.

Together, capital and entrepreneurs are launching a laboratory "revolution." In this article, 36kr tries to answer the following questions:

1. Why can laboratory automation become a tuyere in China? Where is the market ceiling?

2, based on product form and landing scene, which segments have opportunities?

3. Is it worth the investment? What are the characteristics of a good target?

First, the emergence of new scenes and old technologies

Through the statistics of the financing data in the past two years, we can intuitively understand that laboratory automation has become a "visible" outlet, behind the rise of demand in addition to investors in the field of innovative drugs and devices "stumble" after a rush to seek a way out of exploration. 36kr summed it up in the following six points:

① The epidemic is a very direct "catalyst". Not only routine clinical diagnosis requires more rapid and accurate test results, but also pays more attention to personnel safety. The high infectivity and pathogenicity of viruses have put forward the most direct demand for the automation of the whole process involving sample processing, inspection and testing, and sample storage. The automation demand in the field of life science has been verified again, and the rigid demand for the automation of relevant staff has become irreversible.

② With the rapid development of downstream applications in life sciences, the laboratory field is also facing great challenges, such as drug research and development and screening, gene sequencing and other popular technologies for massive real-time data and advanced technical methods to increase the demand.

At present, the domestic substitution window period of the life science support industry represented by laboratory instruments. Due to the characteristics of the industry, the customer stickness of the life science support industry is high, and local pharmaceutical companies have long relied on imported products - but the delivery time is often several months, some products are even out of stock, and the price reduction pressure brought by medical insurance negotiations and collection. In order to ensure the safety and stability of the supply chain and low cost, the domestic substitution of upstream products is an inevitable trend.

(4) Because this subdivision track is quite interdisciplinary and has a high threshold for entry, it can form a monopoly on technology or patents, and it can generate relatively stable cash flow at an early stage. At present, the awareness of risk aversion in the whole venture capital circle is enhanced and the desire for "revenue generating" is extremely strong, laboratory automation has become everyone's "heart".

In view of the fact that automation has been relatively mature in industrial, 3C and other fields, we are eager to find landing opportunities in new vertical fields, because the degree of automation is low and the industry ceiling and high added value, the life science field has become a new track that is jointly locked.

⑥ From an incremental point of view, in order to improve the rapid response ability to major epidemics, China has ushered in a boom in the construction of biosafety P2 and P3 laboratories, or will drive tens of thousands of large-scale laboratories to invest in the construction, for laboratory automation and intelligent manufacturers, which is also a huge business opportunity.

Overview of investment and financing in the field of laboratory automation in the past two years (based on publicly available information, 36 kr mapping)

Driving force: Calculate the "economic account"

From the perspective of the environment, "before the money is very good to earn, now the money is not good to earn", enterprises in particular need to reduce expenses, a key measure is to carry out intelligent transformation, but the premise is to calculate the "economic account" : That is, after the downstream customers complete the automation transformation, the flux brought by the unit space and manpower expenditure is much higher than that of the traditional way, and there is still significant investment income after deducting the purchase income.

Approximate costing for traditional VS automated methods (36 kr drawing)

At present, laboratory automation is a "stock change + incremental rise" business. On the one hand, the form of automation is constantly upgrading and needs to be updated, and it is mainly used in drug research and development, clinical testing, genomics solutions and other fields - the market share of the above is only about 10%; On the other hand, some new scenarios are being mined and applied to form an incremental market involving life science research, molecular diagnostics, synthetic biology, cell culture, etc.

Therefore, at present, for the potential scale of laboratory automation, the market has no authority and unified calculation caliber. Based on interviews with investors and entrepreneurs, as well as backtracking on market demand, 36kr summarizes three main measurement methods, all pointing to the 100 billion market:

First of all, there are nearly 13,000 laboratories in China (including universities and research institutes, hospitals and third-party medical testing centers, new drug R&D enterprises, CRO companies, etc.), with a total cost of RMB 10 million per unit, or RMB 120 billion. In addition, the domestic laboratory analytical instruments market size of 70 billion, if the full range of automation transformation and upgrading, will also bring 10 billion market.

Secondly, the potential market space of China's life science research market is nearly 650 billion yuan. Referring to other mature industries, automation technology is expected to bring 20% incremental output value to the market through quality efficiency improvement, that is, 130 billion yuan; If coupled with the full replacement of the nearly 50 billion yuan market size of traditional human research services, this will be an unmet market with nearly 200 billion yuan of potential demand driven by downstream industrial applications.

Finally, because of the particularity of the laboratory automation industry, it will involve scientific instruments, reagent consumables, automation equipment, information software and other different formats in the process of commercial landing, and are in a stage of rapid growth, in total, the global market size of nearly 200 billion US dollars.

However, because the market is still in the early stage of development and lacks typical benchmarking cases, large-scale application is still out of the question. Therefore, some investors jokingly: "There is no point in talking about scale now, and it is practical to run several customers and open the market."

Evolution logic: The desire for a "black light lab"

Although it has been the attention of the venture capital circle that is used to "chasing the new", laboratory automation is not a new thing, and because of its rich connotation and diversity of scenes, a unified concept has not yet been formed.

In general, laboratory automation refers to the use of a variety of automatic testing instruments and computers and other means to achieve experiment and data processing, so as to reduce the manual operation of experimental personnel and improve work efficiency. The automatic nucleic acid detection equipment and automatic distribution system spawned by the epidemic are typical. Technology is advancing, functions are improving, application scenarios are expanding, and laboratory automation has become a dynamic concept that is iterating at any time.

Maggar technology drawing

Laboratory automation originated in the field of clinical testing and was proposed by Japan in 1981. In the 1990s, commercialized laboratory automation systems entered the labor-scarce markets of Japan, Europe and the United States, and cultivated several well-known brands in the industry since then, such as Hitachi, Beckman, Decon, etc. These brand equipment followed into the Chinese market and slowly rolled out, and local brands also rose later.

Behind the development of the industry is also accompanied by the mature evolution of the form of laboratory automation, which has brought about the improvement of detection flux, the extension of the scene, the gradual opening of the experimental closed loop, the improvement of flexibility and ease of use, and the gradual reduction of unit cost.

At present, according to the degree and scale of automation, laboratory automation can be roughly divided into four levels: single-module form automation, workstation form automation, assembly line form automation, and robot form intelligent. They are not purely full alternative evolution relationships, but rather match different product formats based on customer profiles for cost needs, throughput needs, research and clinical needs. From the functional point of view, laboratory automation as a whole is evolving from the direction of assisting people to replacing people.

In China, the degree of automation in most laboratories is still mainly in the form of single modules; Only some laboratories in the field of workstation form, the degree of integration is limited, the price in the domestic market is mostly at the level of millions of yuan, customers include pharmaceutical companies, CDC, third-party testing centers, etc., representative companies are Thermo Field, Beckman, Agilent, Deken, Hamilton, etc.

In addition, there are parts of the pipeline form automation (Total laboratory automation, referred to as TLA), which through the guide rail series adjacent functional areas, to achieve unified quality control of the whole system, can effectively reduce the human error and biological contamination rate in the experimental process, common in the laboratory - mainly biochemical immunity, The representative enterprises are Beckman, Roche, Siemens, etc., and the local manufacturers Antu biological and the above companies are fighting seriously.

Robot form intelligence is still in the early stage of development in China, has not formed large-scale application, and there is not much gap at home and abroad.

Eventually, everyone's vision points to an unattended "black light lab," in which researchers use the platform to program their ideas, and the system can be efficiently completed. However, this often needs to meet the following requirements: the introduction of advanced technologies such as manipulators, machine vision, and AI decision models; Minimal transformation of laboratory and key inspection equipment to achieve system cascade and closed-loop opening; Instead of manual completion of repetitive work or human can not reach, process result reading and judgment, automatic trial and error and final result screening output.

At present, Eli Lilly Pharmaceutical, InSI intelligence are trying in this regard. It is reported that Lilly has invested more than $300 million in the use of robotic cloud platforms to improve biological capabilities and automate chemical synthesis cycles; Insili is spending tens of millions of dollars to apply its proven AI platform to robotic lab scenarios, further enhancing its ability to automate drug discovery through a combination of "brain (AI)" + "hand (robot)."

Magnesium laboratory automation system

At this stage, most startups are making efforts in the direction of intelligent robot form, typical such as magnesium technology, Huixiang technology, etc. Most players in the industry have not generally reached the stage of large-scale mass production; Another category is the development of automated applications in certain niche scenarios, such as BDA Intelligent manufacturing (molecular diagnostics), BDA Biological (biological cryopreservation), Innoville (cell culture), and Chengyuan Technology (synthetic biology); Some of them are TLA's domestic alternative or fully open TLA systems, such as Antu Bio, Mindray Medical, Mike Bio, or Sanomide, Ritech and so on; There is also a class of automation software applications, such as Bright Wisdom, Zebrafish, which mainly addresses the need for paperless laboratory scenarios, and ultimately to ensure compliance and data security.

Among them, the first category carries the expectation of unmanned intelligent laboratories - most of which are still in the early stages of exploration. As mentioned above, the demand has reached an unprecedented stage, coupled with the extensive exploration of the third generation automation technology based on "AI+ robot", combined with equipment, LAS and LIS/LIMS, life science laboratory automation has become another arena for everyone to fight technology, fight scenarios, and finally scale.

The second type of development is mainly based on the rapid development of downstream industries. Take molecular diagnostics as an example, it is currently the most technically demanding, fastest growing and most challenging branch in the field of in vitro diagnostics, but because of the advantages of high throughput, high sensitivity and relatively low unit cost, with the development of PCR, NGS and other technologies, its application range has been rapidly expanded, putting higher requirements on upstream services.

The third category takes advantage of the development of the county medical market and the normalization of nucleic acid testing. The outbreak and continuation of the novel coronavirus epidemic, heavy testing work and time-consuming testing time, the presence of false Yang (Yin) and possible infectivity in the laboratory, make the automation of testing laboratories gradually become the industry's buzzword; In addition, the contradiction between the growth of the primary medical inspection market and the mismatch of manpower supply brought about by the centralization of county medical care has further promoted the development of the industry.

According to the director of a regional laboratory of Golden Field Medical Inspection Group, there are currently 1,500 medical inspection laboratory automation lines in the country, the overall market capacity of about 7,000, and now the annual growth rate of more than 300 installed machines, coupled with the general price sensitivity of grassroots service providers, which gives local manufacturers greater opportunities. However, this market is very competitive, with predictable ceilings, and is not widely concerned by the capital market, so this article will not discuss it.

Second, when local manufacturers break through

Strictly speaking, the domestic market is still in the early stage of development, so many different forms of products and multiple landing scenes, where to find a breakthrough? Three basic requirements must be met: first, a full understanding of the requirements and results of automation from the goal; Secondly, the technology can be realized, and the process is relatively fixed, meeting the premise of landing; Finally, the system/product is mature and reliable, and customers pay for it, forming a commercial closed loop.

Result-oriented: replicating experimental results and exploring new mechanisms

Results-oriented, we first need to know where the demand for laboratory automation is reflected. If you refer to industrial automation, the key is to reduce costs and improve efficiency, and the final result is a processed or assembled product.

In contrast, laboratory automation ultimately produces results that can be divided into two categories: first, the completion of repetitive actions to produce an accurate data, which is mainly for detection laboratories; The second is to obtain data through a designed experimental process, explore and understand new mechanisms, or screen target compounds, which is mainly for R&D laboratories.

The operation steps and processes involved in the above two processes are mainly for the repetition of the experiment. At present, the application of the detection scenario is relatively mature, and the research and development scenario often needs to use the experimental data to analyze the experimental steps and the specific details of the experiment, so as to optimize the experiment. From this point of view, laboratory automation is not only the operation of the action execution level, but also includes the sample flow, information flow and operation flow. In addition to the operation of the experimental instrument, it is also necessary to integrate the results of the experimental instrument or the data results of the experimental process.

The realization of the above functions generally needs to meet four requirements: AI engines, mobile platforms, multi-modal sensors, etc., to achieve a full range of data collection in the laboratory; Access third-party instruments through the central control system to optimize the working mode; The data of the whole process of the experiment were collected by visual sensors, mechanical sensors, etc., and the experimental process was continuously optimized by steps that were not normally observed or analyzed. AGV robots are introduced to ensure flexibility and connect physically more segmented "independent functional areas".

Landing premise: technology can be realized, to process

For the automated exploration of testing laboratories, its technical threshold is slightly lower, the current industry has relatively large-scale or mature applications, mainly for independent testing, clinical diagnosis of standard projects, and nucleic acid testing due to the epidemic "multiplied ten times", they generally have a large market size, but most of these areas have become competitive Red Sea.

For the exploration of research and development laboratories, entrepreneurs generally reflect that in fact, the above technology is not difficult to separate, difficult to integrate it, achieve full automation, and well cope with the "complexity" and "non-standard" of life science scenes - these are high flexibility and precision requirements, as well as the rise of demand in some emerging technology fields.

Typical new technologies include gene editing technology, IPS cell culture technology, etc., which will bring great changes to human health and life. In this process, automation and intelligence will greatly promote the development of the industry. Taking the gene sequencing industry as an example, automated sequencing technology throughput is now more than 10 million times higher than it was 20 years ago, and its development rate is far faster than Moore's Law in the semiconductor industry. Similarly, the premise of scale application in some emerging fields is also that the automation technology involved in many links has been broken through, so that things that were very difficult to do in the past can now be solved by processes and standardized means.

For example, Alphafold2's cracking of the protein molecular folding problem for amino acid sequence prediction, the progress of AI in crystal type prediction, and the scale application of mRNA therapy technology have all turned biological problems into computational problems. As well as the evolution of semiconductor technology, the computing power breakthrough has opened up the path of computing power to the research and development of new biological drugs. If these scattered technologies can be gradually connected, the entire mass trial and error behavior of biomedicine can be turned into repeated work by machines.

Taking the automated development of macromolecular drugs as an example, at present, deep learning has a significant effect on the improvement of protein structure prediction, which can accurately predict the rough shape of most proteins in three-dimensional space, and help biotechnologists identify and produce proteins, so as to make the research and development of innovative macromolecular drugs predictable and programmable, and improve the efficiency of the whole process of drug research and development. However, most of the subdivision directions are still in the early stage of exploration, and there are fewer actual landing cases.

Commercialization: Find buyers and form benchmarking cases

After the application of technology, finding benchmark customers is the last step to open up the closed loop of business.

Ideal is very plump, reality is very bony. Due to the high cost of overall laboratory automation construction (dominated by imports), customers with both rigid demand and payment ability are mainly concentrated in large CRO, CDMO companies, pharmaceutical companies, laboratory departments and third-party medical laboratories of large hospitals, as well as some government-funded benchmarked laboratory projects. But these customers generally lack interest in local brands.

A biological drug CDMO company innovative biological drug research and development and pharmaceutical center IT director told 36kR, "At present, because of the domestic robot equipment dead corner design, in addition to domestic alternative and much cheaper, can not find other reasons to use, not to mention the big factory is not poor money." However, he also said, "If it does not involve the core GMP part (some equipment with high precision requirements), it should be localized, and it will slowly penetrate."

It is said that in some non-high-end technology fields, Chinese brands already have significant advantages, such as automated pipette workstations, biological sample storage, detection probes, etc., which are not only much cheaper, but also easy to use; However, in some high-end automation equipment involving biology, computers, medicine, machinery industry and other fields of patents and technology, there are some upstream parts have not yet achieved localization, domestic brands are in a "stuck neck" state, the future in the supply chain or face challenges, will become an important card point for domestic brands to break through.

In addition, a necessary condition for the production line to move from the research and development stage to the laboratory application is that the production line is standardized and replicable. However, in the field of life science laboratory automation, customers are small, scattered and immature, "non-standard, small batch, multi-variety", "can not raise demand" has become a "common disease" in the industry. A startup company said that a customer had bluntly said: "I don't know what my needs are, your team will first stay in the field, talk to all departments and all links of people, to help us transform it."

To this end, local laboratory automation manufacturers often have to play the role of customer program planning and industry leaders, and need to polish products through benchmarking customers/cases to prepare for subsequent large-scale applications. A typical Youmeijia technology built its own Kunpeng laboratory and gradually accumulated benchmark customers through exploring cutting-edge research in life science. When asked why they want to build their own laboratory, MGI Technology CEO Huang Yuqing said, "to build such a laboratory investment is very big, the operation is very difficult, and the cycle of looking for customer cooperation is too long, it is better to do it themselves."

Huixiang Technology is cooperating with AI pharmaceutical company Insi Intelligence to build an intelligent robot drug research and development laboratory to explore fully automated applications in unmanned scenarios. It is reported that six companies of the same type came to bid for the order, and peers evaluated, "Even if you have to accept the order at a loss, the investment in the early stage is worth it."

And for what is really a good product, we also have a basic consensus: that is, the production line can make money, the model can be copied, "others can't do it" (IP), the subsequent cost is reduced, the deployment cycle is shorter, and it can be expanded to other types of customers, changing the market pattern. In this regard, a typical case is the Falcon series, a small flexible intelligent delivery system developed by Novo.

Three, is it worth it?

Early commercialization is difficult for every entrepreneur to face the problem, investors are also betting on a "life science laboratory intelligent" future. And when will it come?

The next 5-10 years will be concentrated

Liu Wei, managing partner of Zhang Ke Lingyi Venture Capital, said that the CRO industry can be likened to the development of this industry will be divided into three cycles: the first wave, the industry from scratch, "there are companies dedicated to do this thing", as in recent years, Xuanjian Technology, Benyao Technology and other companies focused on robot intelligent products have emerged; In the second wave, a group of companies began to be willing to pay, constantly run in trial and error, some companies to develop and grow, the market began to reshuffle; Finally, find out the demands of the first batch of companies, solve the universal problem, and then provide standard products, rapid replication.

"The first wave has just begun, and the opportunity is the biggest, and the next five to 10 years will be the stage of acceleration and even concentrated outbreak, when the market will have universal, cost-effective and independent intellectual property products." Moreover, for investors, the valuation of the upstream of the biomedical industry is also relatively reasonable, and it is an investment theme given by The Times.

According to the actual business attributes of the future company, Yang Xiaolong, partner of Innovation Works, pointed out that several types of industries can be evaluated: first, if it can become a labor-based company, the valuation can be compared to CRO companies such as Wuxi Apptec; Second, it can produce biological results, such as specific drugs or compounds, which have corresponding asset value, and can be valued for standard drug enterprises, generally using discounted cash flow method; The third is automation equipment, which can be used to value overseas listed companies such as Beckman, PE and Agilent.

However, he also stressed that given that this is a highly interdisciplinary field, in fact, it can also be valued with a comprehensive reference to the above three industries, which can produce a certain advantage premium and match the development potential of the industry.

The look of a "good target"

In the investment market, what is a good target? Yang Xiaolong believes that it is either a system driven by super key components, or it is to train it into a "brain" - that is, a system of human judgment, which can be boiled down to three elements: a "card neck" point, with independent intellectual property rights, and a special structure.

From the perspective of the industry chain, like other industries, downstream customers generally pursue cost-effective and "user-centered" solutions, and multinational manufacturers are generally difficult to meet the needs, which leaves opportunities for local manufacturers. However, the disadvantage of this model is that it is difficult to modularize and standardize - this is the biggest gap between niche market and blue ocean market, and the market valuation is often lower than standard products. For automation developers, the key is to strike a balance: the protocol developed for a specific project is gradually optimized into a standard module, through the assembly of standardized modules to meet customized needs, and fundamentally reduce research and development costs.

As for the landing form, when asked whether to choose software or hardware, most entrepreneurs, investors and even customers will vote for hardware or software and hardware integration.

Drawing on the successful overseas experience, Dassault BIOVIA (small molecule medicine) and Benchling (large molecule medicine), as the representative companies, entered the laboratory data recording market through self-developed software, the latter has been valued at 6 billion US dollars, and it can also form a business barrier through the asset-light software system. However, people in the industry generally believe that infrastructure is the entrance of life application science, and equipment is the carrier of data formation; Software tools alone cannot expand the market, especially in the domestic market.

One investor told 36kr that benchling was valuable because it began as a differentiated digital tool and gradually expanded into a communication tool for hundreds of thousands of professionals in the scientific community and industry, which deposited more and more templates and databases, "This is very valuable." However, in China, doing such tools will "roll" badly, not only to fight price, fight service, fight customer sentiment, Party A is also used to customization, most manufacturers may not go to PLG (Product-Led Growth, product-driven growth) "hang up".

The apparent barrier

The barriers of this format are mainly reflected in technology and team, customer reputation and industry awareness and accumulation.

First of all, laboratory automation is also regarded as a life science support industry, involving biology, chemistry, materials, electronics and machinery and other professional fields, the products of various subdivisions of the industry often need to be applied in multiple disciplines, usually requires a composite team for years of technology accumulation to develop and manufacture mature products.

Secondly, there is customer choice inertia. In general, downstream customers are very cautious about the choice of products, tend to purchase high brand recognition, good market reputation tool products, once locked suppliers are difficult to replace.

Finally, in the long run, in addition to "technical barriers", the real stable moat also comes from industry cognition, cost advantage and product matrix, which requires enterprises to take root in the scene, constantly accumulate data and polish products, and quickly start volume, and then have the opportunity to achieve cost advantages; In addition, in order to resist the cyclical risks of downstream customers, the company should also choose a good track and constantly improve the product matrix.

As China joins ICH, we need to look at the industry from a global perspective, and "Why China" has become an important topic - in the global market with a higher degree of marketization, the fields that give full play to the advantages of China's intelligent infrastructure will rise at a faster pace, including laboratory automation. Under the cultivation of the global market, future laboratory intelligence will also become a reality sooner.