How can the mobile robot industry, which is popular but not profitable, break through the current predicament?

In 2008, the robot WALL-E became the most popular robot idol. As a tracked garbage sorter, it worked alone in the apocalyptic wasteland for hundreds of years, and had no complaints about the long working hours and dangerous working environment. In real-life scenarios, mobile robots are increasingly replacing humans in heavy, repetitive or dangerous and harsh work environments, helping humans reduce labor intensity, save costs, and improve happiness.

Mobile robots are a highly anticipated development direction in the industry in recent years, but behind the hype there are also real problems such as being well received but not popular, fierce price wars, and focusing on sales rather than technology. This article attempts to explore the following questions: how can the mobile robot industry break through the current dilemma, what potential innovation opportunities do startups have, and what are the viable business models for the future.

Core Viewpoint

1. The industry is full of expectations for the development prospects of mobile robots, but the actual implementation speed is still lower than expected, and it has fallen into a dilemma of being well received but not popular.

2. Looking to the future, there are three potential innovation paths for mobile robot companies: continuous improvement of mobility capabilities; moving from single-point intelligence to global intelligence; and the integration of mobility + execution capabilities, which is expected to help mobile robots break through the existing market ceiling.

3. With the advancement of technology, the business model of mobile robots will also undergo profound changes in the future. Customers may not buy simple products, but scenario-oriented overall solutions or capability modules.

▲ Image source: PIxabay (based on CC0 agreement)

The dilemma of mobile robots: popular but not profitable, fierce competition, strong sales but weak technology

Mobile robots are an important branch of the development of the intelligent robot field. In the early stages of the development of mobile capabilities, in order to meet the needs of large-scale and heavy material handling and turnover in the industrial and logistics fields, mobile robots effectively exerted basic mobile capabilities in fixed scenes and fixed tracks through various methods such as track traction and magnetic strip navigation. With the gradual maturity of SLAM technology, robots can move in semi-open scenes and flexible tracks. A new generation of mobile robots represented by AMR and unmanned forklifts began to take the stage and began to be applied in more scenarios such as logistics and commercial services.

这些年来，行业对移动机器人的发展满怀期待，巨大潜在市场空间也陆续吸引了大量参与者入局。但现实是移动机器人在各场景的出货量还没有迎来井喷式爆发，也没有公司真正在行业内占据显著领先地位。目前移动机器人领域呈现出叫好不叫座、同质化价格战和强销售轻技术的困局。

Mobile Robot Innovation Path

The future of mobile robots is bright, but how can we break through the current dilemma? This is a key question that many startups and entrepreneurs are thinking about and exploring. We believe there are three potential innovation paths.

1. Continuous improvement and breakthrough of mobile capabilities

对于任何种类的机器人来说，能够实现大规模应用的前提一定是显著的价值创造和效率提升。这体现在机器人比人类在工作效率（同样工作比人做得更快更持久）、工作能力（更好的力量速度精度），或是工作耐受性（做危害人健康的工作）等多方面的优势。

The above theory also applies to mobile robots, but the main bottleneck is the level of intelligence. The current level of intelligence of mobile robots can be compared to the level of L3-L3.5 under the framework of autonomous driving. If we can combine the existing technical level, find the most suitable landing scenario, and fully reflect the efficiency advantage of the robot solution, commercial success may reach 50%.

在酒店、医院、餐饮为代表的商业服务领域，移动机器人能力和需求已经有了较好的结合。但在环境更复杂、节拍要求更高的场景中，机器人的智能水平尚不能满足产业实际需求。比如在物流、工业场景中，人的工作效率仍然显著高于移动机器人。为了满足工厂的实际装卸货需求，我们经常能够看到叉车工人在物流仓库里施展高速过弯、漂移的神奇操作，而机器人的智能化水平尚不能与工人的熟练操作相媲美。有限的价值创造，也使客户很难为目前的移动机器人产品慷慨买单，这造成了工业客户从试点、小批量到大批量应用机器人的漫长商业周期。

The implementation of industry applications should be comprehensively considered from the perspectives of technology + product capabilities (supply) and industry challenges + cost structure (demand). We believe that with the improvement of technical capabilities and the evolution of labor cost structure, mobile robots will have the opportunity to be implemented in more complex and high-performance scenarios such as industry in the near future. Startups should not only focus on short-term market competition, but should look to the long term and look for technology research and development and innovation that can improve the intelligent capabilities of mobile robots, including perception and decision-making (VSLAM, semantic recognition) and other aspects.

In addition to improving mobility, the form of mobile robots also needs to continue to evolve. The wheeled form currently used by mainstream mobile robots may not be the optimal form for achieving universal mobility. Although the bipedal/multi-legged solutions represented by Boston Dynamics are still in the early stages of technological development, they will have better adaptability to complex environments such as outdoor delivery and security inspections.

▲ Image source: MiR and Boston Dynamics official website, top: autonomous mobile robot; bottom: Boston Dynamics' quadruped Spot and bipedal Atlas

2. From single-point intelligence to global intelligence

In many scenarios, mobile robots are an important part of production/business processes. This means that while ensuring their own mobility, mobile robots need to cooperate with the business logic and processes of the actual production environment to make direct contributions to core indicators such as production efficiency and service quality that customers actually care about.

对工厂来说，智能工厂的定义不仅仅局限于物流自动化这一个范畴。半导体和PCB领域是很好的代表——作为离散生产中最接近流程的行业，半导体在设备自动化和工厂信息化领域，已显著领先于其他离散行业。某头部客户把工厂智能化的定义拆分成以下几个维度：

Logistics automation level

Improved production efficiency

Product yield improvement

Assist intelligent decision-making capabilities

在半导体生产流程中，通过人或者移动机器人完成物料的及时转送是最基本的要求。如何能够有效配合／协调各道工序之间的生产节拍，避免呆料和设备排队等行为的发生，是显著提高生产效率甚至良率的关键。这就对机器人厂商提出了更高的要求。

In the long run, industrial production and delivery models will undergo profound changes. In industries such as clothing, retail, and electronics, the change in demand for small batches and large batches will drive factories to change their existing production and collaboration models to make them more flexible and flexible. This will also drive factories to make greater use of software capabilities to achieve more real-time and efficient production scheduling, optimize material turnover and production logic, and continuously improve production efficiency and quality.

In such a production environment, the hardware of mobile robots is only a handling device to complete material turnover. The new generation of startups strives to become the "on-site brain" in the future industrial production environment, deeply integrating the process and production logic of customer business, connecting multiple systems such as MES, WMS, WCS, and thus realizing global and full-cycle intelligent scheduling capabilities.

In addition to industry, we believe that similar opportunities will emerge in the commercial service sector. In the near future, machine intelligence is expected to open up the entire business process of hotels and restaurants and become a smart operator in different scenarios.

3. Deep integration of execution and mobility capabilities to create greater value

There is a view in the industry that the long return on investment (ROI) period for customers prevents large-scale deployment of mobile robots. If the cost of purchasing and deploying robots can be reduced and the ROI period can be shortened, customers will be more motivated to try or use robot solutions.

We believe that we should try to shorten the customer's investment return cycle from two aspects: reducing costs and increasing output. How to make robots play a more important role in the scene, replace more traditional manual labor, and create greater value/output is the direction that technology innovators should pay more attention to.

The current mobile robots can only perform simple handling and turnover tasks in industrial and logistics scenarios, and due to the limited driving speed, robots only replace 50%-60% of workers' duties. From the customer's perspective, limited value creation may be a more important factor hindering procurement.

In future work scenarios, robots are expected to simultaneously move (feet) and perform (hands), thus becoming composite robots. The concept of composite robots is not new, but it will take time for them to mature.

At present, it is common for robot manufacturers to purchase products produced independently of each other, and then DIY assembly and secondary development. This integration model makes it difficult to control the cost of composite robots and also leads to a more complex structure of composite robots. The controllers of each sub-unit are independent of each other and difficult to control as a whole, which limits the flexibility and scalability of robot control and functions. In the future, we believe that the mobility and execution capabilities of composite robots are expected to achieve deep integration, achieving the effect of 1+1>2.

▲ Image source: Robot Center, the common forms of composite robots at present; the ultimate form of composite robots in the future is still unknown

Future-oriented business model

product

The lack of mobile robot capabilities has caused current startups to spend a lot of energy on delivery, implementation, and maintenance. With the improvement of robot intelligence, future mobile robots will have the opportunity to be fully commercialized in both industrial and commercial service scenarios, allowing customers to experience a consumer-level experience that is ready to use, ready to go, and long-lasting and reliable.

Serve

In the second path mentioned above, the factory brain or service center that combines software and hardware is expected to become a carrier that perfectly replaces certain functions or scenarios. In the future, customers will not pay for simple robot products, but for more efficient intelligent industrial and commercial capabilities. This is the true meaning of Robot as a Service.