DragonFly: The People’s Autonomous Vehicle

autonomous vehicles for citizens globally who are potentially run the risk of being excluded from the pending transportation revolution.

The benefits of autonomous driving are often touted and well known: they can cut carbon emissions, reduce accidents and deaths and create more efficient transportation economies in cities around the world. However, to truly scale the benefits of autonomous driving to citizens globally and not just the privileged few, we need an innovative approach towards lowering the cost of autonomous vehicles while remaining uncompromising on safety.

Currently, the ubiquitous adoption of clean energy autonomous vehicles (AV) has several barriers such as technological reliability, ethical and legal concerns and affordability. To address these issues, we have developed an autonomous vehicle for under $10,000 called the DragonFly equipped for low-speed scenarios such as university campuses, industrial parks, farms and areas with limited traffic. Our approach starts with a low-speed version of the DragonFly to ensure safety thus allowing immediate deployment. Following technology improvements and feedback, we will be able to deploy vehicles suited for high-speed scenarios. Ultimately, we aim to have our vehicle’s performance equal that of a human driver in any type of driving scenario.

To achieve Universal Accessibility of Autonomous Driving, we have developed very easy to use interfaces for people with no prior technology background to manage and control autonomous vehicles.  In addition, we have collaborated with universities worldwide, as well as the IEEE, to develop courses to equip workers with the necessary skills to manage and control various kinds of autonomous vehicles, such as delivery robots, industrial robots, farming robots, and other AV applications.

Once scaled at a global level, citizens most affected by technology-driven transformations can reap the benefits of the DragonFly at a personal level to create increased independence, freedom, affordability and upward economic mobility in their lives. An obvious but nonetheless significant benefit for citizens in utilizing the DragonFly includes taking back time for leisure and enhanced productivity from participation in an increasingly demanding global information economy.  According to the The Boston Consulting Group (BCG), Americans can recover a total of 30 billion hours per year that they now spend driving, sitting in traffic or looking for a parking space.

There is also a true opportunity cost to driving, parking and insurance beyond increased time savings and enablement of upward economic mobility. A McKinsey report states that from 2015 to 2030, the benefits of AVs could add between $3 billion and $4 billion, or $2,200 to $2,800 per resident. Nearly four-fifths of these benefits will result from improvements in safety. The report also mentions that a developing, dense city of average size could realize $600 million in annual societal benefits by 2030. Specifically, in the US alone, the universal adoption of ride-sharing and clean-energy autonomous vehicles could save as much as $520B annually. This figure almost ties with the GDP of Sweden, one of the world’s largest economies.

Finally, AVs can offer freedom and independence to people who cannot drive cars: the very old, the very young and the disabled and enable more diverse participation in our social and cultural networks.

DragonFly Car is a safe and reliable low-speed autonomous vehicle with total cost of less than $10,000 USD that can be scaled globally.  The first innovation is the patented vision-based sensor fusion: fusing multiple affordable sensors synergistically to generate accurate real-time perception and localization results. These sensors complement each other such that when one fails or otherwise malfunctions, others can immediately take over to ensure system reliability. The second innovation is the patented modular computing system design which enables each module to perform more computing. The result is a reduction in the burden on the main computing system.

With over 120 international patents around our core technologies, we have integrated numerous technologies into this product, including visual sensing, localization, perception, planning and control, and system integration. Please see more details below:

Computer Architectures for Autonomous Driving (https://ieeexplore.ieee.org/document/7999133/)

A Unified Cloud Platform for Autonomous Driving

(https://ieeexplore.ieee.org/abstract/document/8220475/)

Creating autonomous vehicle systems (https://www.oreilly.com/ideas/creating-autonomous-vehicle-systems)

Creating Autonomous Vehicle Systems
Synthesis Lectures on Computer Science

(https://www.morganclaypool.com/doi/abs/10.2200/S00787ED1V01Y201707CSL009)

Ubiquitous Adoption: transforming vehicles into autonomous vehicles

We plan to enable many different kinds autonomous vehicles with our DragonFly technologies, and most importantly, allow those without any technology background to easily manage and control different kinds of autonomous vehicles.  

We have enabled cargo carrying vehicles and low-speed autonomous vehicles, the next target is autonomous farming machines. With different kinds of vehicles, we provide a unified interface such that those without any technology background can learn once and apply his/her autonomous vehicle control skills to a wide spectrum of autonomous vehicles. 

Universal Accessibility: Learn Once Use Anywhere

Within the next three to five years, our focus will be on "learn once use anywhere". We provide a unified and easy-to-use interface for those who have never been exposed to technology.  This way we upskill these workers and allow them to learn once but also apply what they learn to different autonomous vehicles. For instance, we can train a farmer to manage an apple-picking farming machine.  Later on, the farmer can easily apply his skills to control delivery vehicles, thus greatly improving his value on the job market. Eventually we reach universal accessibility. 

There are different ways we are helping the robotics and automotive industries with the DragonFly’s technology. First, we are providing the DragonFly sensor as an independent component for robotics and automotive companies to improve their own solutions. Second, we are designing and implementing hardware and software solutions for different usage scenarios that require transportation and automation services, including farming robots and industrial transportation robots etc.

As a first application, we are currently deploying the DragonFly cars within an industrial park, providing short range transportation services for more than 10,000 employees working in the industrial park. These employees have little prior technology background but quickly learned to use our unified user interface to manage and control these autonomous vehicles. 

One snapshot of the UI is below, which shows the current status of the vehicle and allows the user to control the vehicle through interacting with the UI on a mobile device. 

In the next 12 months, we will focus on serving 100,000 people with our DragonFly technology. These pilot customers should be able to share their experiences with DragonFly technology within their communities and help us reach 10 million users within the next three years.  

Our team has an expansive skill set that equips us to integrate multiple technologies into our product and bring this product to market.  Our engineering team consists of over thirty industry experts in artificial intelligence, system integration, hardware engineering, and optics. Our business development team is led by an Oxford MBA with over five years of business development experience.  Since founding PerceptIn, we have successfully attracted over 12 million USD of funding from top venture capital firms including Walden International, Matrix Partners, and Samsung Ventures.

We have a two-layer revenue model:

1.) Hardware Business-to-Business Model: we provide the DragonFly sensors to enable visual perception for various kinds of robots and vehicles built by other robotics and automotive companies.

2.) Solution Business-to-Business Model: we also provide different transportation robots with full-stack hardware/software solutions for the target scenarios specified by our customers.  This model helps us penetrate into different vertical industries, including farming robots, industrial transportation robots etc.

PerceptIn focuses on enabling visual perception for robots ranging from IoT-grade robots to autonomous vehicles. We have just scratched the surface of robotic visual perception capabilities and we can expect this trend to be a major disruptor in the coming years; in fact, there’s no end in sight for this technology as it is set to transform nearly every sector of the economy and improve the lives of citizens globally who may otherwise be at risk for being marginalized by robotics advancements.

I founded PerceptIn in 2016 to solve the robotic perception challenge via an affordable, vision-based, multi-sensor and fusion-based system that utilizes a combination of computer vision technologies. My mission over the next five years is to enable visual perception for as many robots as possible, from IoT-grade robots to higher-end autonomous vehicles. While we have created very advanced technologies to improve human quality of life, we need a better channel to convey our message and to establish more strategic partnerships globally, particularly those in emerging economies.