Exploring the Future of Delivery Bots with Cartken's Christian Bersch

CEO shares thoughts on regulatory environment, computer vision, industry's future

Cartken has racked up some impressive wins in the PDD / delivery robotics space, including partnerships with Grubhub, unique deployments in the hospitality space, and more. Despite this, the company seems to fly a bit beneath the radar - attracting less attention than the likes of Coco, Kiwi and Serve. To help fill in the blanks, we sat down with company CEO Christian Bersch, who has a deep background in robotics from previous stints at Google and Bosch, to learn more.

OttOmate: Tell us a bit about the founding team, and how you all came together to create Cartken?

Christian Bursch: Cartken was launched by a group of former Google, Apple, and Bosch engineers and operators with a long history of building emerging technology products. Our team has deep expertise in self-driving vehicles, AI-powered robotics, and delivery operations.

After multiple years spent on developing and launching emerging technologies, the group came together in August 2019 to build autonomous delivery robots that are more accessible and commercially viable. By applying that knowledge, our team was able to develop a fully integrated system in less than a year. We first launched robots in California.

Company org chart

“We are also working directly with regulators - for example, through our partnership with Mitsubishi Electric in Japan.”

O: What has your collective background in robotics and AI taught you about this space? Are there any particular lessons you think you’ve already learned?

CB: In general, the automotive industry has come to a realization that integrating self-driving vehicles into roads is much harder than anticipated. We are still years away from true robot taxis and delivery trucks, that can drive in all conditions the same as humans can. Also, the cost of developing a self-driving fleet is far from being efficient.

The level of autonomy is very different between companies in this space. Our Cartken robots achieve L4 autonomy, which means they can navigate sidewalks and pedestrian paths on their own within mapped areas. L4 is the highest level of autonomy that is possible for autonomous vehicles - L5 would be autonomous driving without prior knowledge of the driving area, which only humans can do. The other thing to call out is that Cartken’s autonomy stack is based on cameras combined with state-of-the-art computer vision.

Most robots in this space rely much more heavily on or even exclusively on teleoperation. The high level of autonomy of Cartken’s robots doesn’t just lower the cost, but makes their operation much more reliable and safer, as the AI system can react much faster than a human over an LTE connection ever could.

Developing robots that can navigate autonomously around obstacles and pedestrians, but get help from a human when they encounter an unknown situation - is the operation we chose to pursue.

O: As the PDD / robotic delivery market gets busier, how do you foresee Cartken’s approach differing from some of your industry peers?

CB: We think overall, success in the delivery robotics space means having intelligent machines that can navigate among existing environments of people and infrastructure. But, for us specifically, Cartken’s strategy for success focuses on three main areas:

The first is having the most capable and flexible autonomous robot platform on the market. Our technology leverages low-cost commodity hardware and best-in-class sensors and cameras which allows us to achieve high levels of autonomy, safety, and reliability.

The second strategy is the ability to extend our technology to be customizable and brandable. We allow our robots to be branded with our partner’s logo and messaging and want to make sure our partners own their end customer data.

The last strategy for success for us is building a broad distribution network and partner base. We are a technology provider to existing delivery services versus trying to build a new, robot-only delivery service. We offer robots as a monthly subscription. Securing partnerships with a few big players in the space, such as Grubhub, Magna, and Mitsubishi, allows us to tap into their distribution pools in exchange for our technology.

O: Are there particular markets that you see as being most receptive to robotic delivery?

CB: Naturally, restaurant aggregators, such as Grubhub, Uber, DoorDash, and Postmates, were the first to pilot robots as new delivery mechanisms.

We have noticed universities, quick service restaurants, grocery retailers, and ghost kitchens are the primary segments that take advantage of autonomous delivery robots and the infrastructure restaurant aggregators have built. They use robots to streamline curbside pick-up, local delivery, and internal shuttling operations and reduce labor cost. Malls and hotels are catching up with the trend.

Manufacturers, e-commerce companies, and delivery carriers adopt robots for micro-fulfillment operations which allows them to optimize manual processes and save their staff time.

O: Are there any use cases, beyond prepared food, that you see as particularly adapted to robotic delivery?

CB: We have found several use cases for delivery robots operations, and much more to be discovered and tested.

For example, robots can be used for micro-fulfillment in factories and warehouses to shuttle boxes, instruments, and materials between aisles and halls.

Secondly, robots are useful for short-distance shuttling in instances such as:

• Food/medical supplies at elderly care facilities

• Small package delivery in corporate campuses, offices

• Linens at hotels/resorts, hospitals

• Food prep items at stadiums, hospitals, hotels

• Hauling of trash at various types of locations

• Catering for events

Robots can be transformed into vending machines to deliver consumer goods for sale across large designated areas, such as outdoor event venues, conferences, and stadiums.

Last-mile package delivery is not a new concept. A few e-commerce and shipping giants have been developing and testing their own delivery robots, but they haven’t become widely adopted yet. For example, DPD, Europe’s largest parcel carrier, is piloting Cartken robots for last-mile delivery in the UK, and locals receive it quite well.

Lastly, as the technology becomes more mainstream, we predict that consumers will leverage robots to deliver items peer-to-peer (P2P). For example, you can deliver a home tool set to your friend’s house when he/she needs it.

O: How do you feel about the regulatory environment for PDDs? Are you seeing consistent policy frameworks across markets, or is it still a bit of a “wild west” that needs to be made orderly?