Kiwi Campus builds robots to solve the last-mile delivery problem
From Colombia to California
When Felipe Chavez first visited California two and a half years ago he found himself shocked about the high prices of food delivery: Chavez had ordered sushi for $11 and ended up paying another $10 for service and delivery. This was the first time he realized there was a problem. Talking to some friends about what he had noticed, he found out that people in California usually order food about once or twice a month. In contrast, people in his home country Colombia people would usually order multiple times per week. He figured that the difference did not lie in the desire for comfort food, but rather in the difference in prices. In Columbia, Chavez says, labor is much cheaper than in the US, and therefore so are the delivery costs. Originally Chavez had planned to stay in California for only 3 weeks, but he fell in love with Berkeley. Chavez instead decided to stay indefinitely and take on the mission of supporting the evolution of food delivery by making it faster and cheaper.
“I arrived with a tourist visa, planned to stay 3 weeks. I have worked in Tech all my life and Silicon Valley is said to be the best place in the world to start a company. But I fell in love with Berkeley more than with Silicon Valley. Berkeley has a great campus with much talent and an amazing gastronomy offer. I realized that this is where I want to start a company.” Felipe Chavez, CEO and Co-Founder of Kiwi Campus
The challenge of last-mile delivery
The problem Chavez had discovered is commonly known as the “last mile” delivery challenge. “Last mile” is a term used in supply chain management and transportation planning to describe the movement of people and goods from a transportation hub to a final destination in the home. Accordingly, “last mile” might is not always only a mile; it can be as many as 15 miles. This final part of the delivery, despite its comparably short length, accounts for about 30-50% of the total delivery costs. Increasing the efficiency of this part of the delivery, therefore, creates huge potential for e-commerce companies seeking a competitive advantage. A first step in improving efficiency is to build a broad infrastructure of so-called delivery hubs, effectively warehouses. Walmart and Amazon, for example, aim to have warehouses within ten to fifteen miles of their consumers. This leads to the question: did Amazon actually buy a grocery store with its acquisition of Whole Foods or was it just buying the 473 warehouses that Whole Foods had owned and operated?
Now, why is this problem so relevant? First of all, e-commerce has been growing and will continue to grow at a rapid pace. Consequentially, the need for delivery will as well. As of today, the global cost of parcel delivery amounts to around $86 billion. Moreover, consumers’ expectations are steadily rising and same-day delivery is becoming a new standard.
US ecommerce sales, 2014-2021 (historical and projected)
How does Kiwi solve the problem?
When Chavez decided to take on the challenge, he already had profound knowledge about the delivery business and its operational challenges from his previous work experience. Still, he conducted extensive research on the unit economics of delivery, especially “last mile”, and came to the solution that robots were the best solution to the costly problem.
“In some areas humans are underrated, but in delivery, we need automation. We propose humans plus robots. Adding robots makes it more efficient.” Felipe Chavez, CEO and Co-Founder of Kiwi Campus
Compared to drones, which companies such as Amazon are experimenting with, robots have several advantages according to Chavez. First of all, robots are 50x cheaper to produce. They have a better payload capacity and a longer battery life. Additionally, government regulations are currently in favor of robots. Of all these factors, battery life is the main bottleneck. Even if the battery life for drones could be improved though, robots may maintain a competitive advantage as they would benefit proportionally from these improvements. Also, so Chavez says, a key element for success is the integration of his robots into society. While drones are “noisy and annoying” for people, Kiwi Bots have been designed to integrate into civic life and to become a part of society. For this reason, it has been crucial to design them as small as possible and to give them a face. There is still the remaining question of whether this will be broadly accepted by regulators and society.
“I believe that robots are important for society. They will ensure that prices go down, so you’ll need less to live. But also, I believe sidewalks are sacred, so we design our robots so they’re not too big, have a face, have a design, so they can be part of the society instead of a robot doing things.” Felipe Chavez, CEO and Co-Founder of Kiwi Campus
How it works
When I asked Chavez how he gathered the knowledge to be able to build a functioning robot in such a short time, he simply answered “the internet.” This pragmatic attitude allowed him to become one of Berkeley’s most notable companies and to conquer the challenges he and his team faced during the development process. While admitting that developing hardware has been the greatest challenge on Kiwi’s journey so far, this is exactly what drives and motivates him.
“Code is simple to fix, the hardware is another story. If there is one small mistake in a piece of the robot, you have to wait one month to get it fixed. And you need to replace your entire fleet. It is complex, but magical. I believe the best engineering in the world is when you see it that you think that it’s magic. And that’s why I left to work with robots.” Felipe Chavez, CEO and Co-Founder of Kiwi Campus
Another technical challenge lies in the robots’ logistic efficiency. While sending a robot from A to B is relatively simple, navigating and coordinating an entire fleet is a different challenge.
For users, ordering food is very easy: one simply needs to download the Kiwi Campus app, select the items to be delivered and checkout via credit card or Venmo. Delivery is currently priced at $2.80. Food is often discounted due to partnerships, for example with Soylent or Redbull, perfectly targeted to the students’ needs. The Kiwi robot then picks up the food and navigates its way to the consumer using computer vision to recognize objects.
Video highlighting Kiwi Robot delivery cam
Funding, scale, and outlook
To this point, Kiwi has delivered more meals with robots in the world than anyone else (i.e. 9,000 deliveries in eight months). The company operates the biggest active fleet in a single location and is diversifying business by the recent addition of two products: a restaurant bot and a self-driving bike. While success for its investors primarily means generating a high valuation, Chavez is committed to the mission of contributing to society by ensuring cheaper delivery. Kiwi has raised $1.1 million in its seed round and is currently raising its Series A. Before scaling to further locations, Kiwi wants to make sure that its product is working. Currently, the Kiwi fleet consists of 75 robots that are already 50% cheaper than human carriers and need an average of 27 minutes per delivery, which is 8 minutes cheaper. The vision is to ensure $1 / <1-hour delivery everywhere. Kiwi also recently introduced two new products that Chavez and his team will continue working on: a Restaurant Bot and a self-driving bike called KiwiTriike.
Kiwi Campus new product demo