Your Next Airport Meal May Be Delivered By Robot

Robot delivery has long been touted as a game-changing technology of the future. However, it still hasn’t cracked the big time. Drones still aren’t airdropping packages into our gutters by accident, nor are our pizzas brought to us via self-driving cars.

That’s not to say that able minds aren’t working on the problem. In one case, a group of engineers are working ton a robot that will handle the crucial duty of delivering food to hungry flyers at the airport.

Tacos To Gate 37, And Step On It!

Eating at the airport can be a bit of a crapshoot. Seating is usually limited, with thousands upon thousands of people passing through in any given day. Even if you get to the airport early, you might struggle to find somewhere to sit down and eat. Get there late, or get a bad security line, and you might not have time to order food before you need to walk to your gate. Airports are typically large, sprawling complexes, and the food outlet you desire may be on the opposite side of the building to where you’re boarding your plane.

Robot delivery could potentially solve these issues by delivering food directly to passengers at the gate. It’s this problem that robotics startup Ottonomy set out to solve, working in partnership with Cincinnati/Northern Kentucky Airport.

In the cluttered, indoor environment of an airport, GPS and other satellite constellations simply aren’t usable for navigation. Instead, Ottonomy’s Ottobot relies on lidar, cameras, and ultrasonic sensors to detect obstacles and find its way around the environment. The company created proprietary software for mapping indoor environments, in order to give its robots the necessary 3D map to get around their workspace.

Airports can get incredibly crowded, and early on, the team found that navigation in these situations was difficult. This led to the Ottobot’s current configuration, in which every wheel is motorized and capable of steering. This lets the robot crab sideways, execute a zero-radius turn, or tightly swerve around corners, both of which are useful to get around tight spaces. The drivetrain was inspired by electric wheelchairs. Just like an electric wheelchair, the Ottobot needs a tight turning circle and the ability to deal with lumps, bumps and kerbs, at times.

Several compartments on the bot are used to store various items a customer might order. As is becoming common in food delivery, this would allow the necessary separation of cold drinks and hot food, for example. The Ottobot is also capable of raising and lowering its ride height. It might seem like a curious feature, but it helps the robot serve a greater range of customers. The robot can lower itself down to allow a child to pickup an order, or raise itself to help a customer that can’t easily bend over. Notably, the company has kept the cabin design modular so that it can be customized by end-users to suit different delivery roles.

As with most any autonomous system humans have ever built, there’s always a risk that things will go wrong. In the event that an Ottobot can’t figure out where it is or where it’s going, staff can take over control to get the robot out of trouble. This feature is particularly useful if the robot finds itself in an unexpected situation. The robots also feature a “How Am I Driving?” sticker on the back, that invites feedback on the robots performance.

Having a “How am I driving? sticker on the robots is good for accountability, and can help ease public concerns around potential safety issues. Credit: Screenshot, Ottonomy IO – YouTube

The order process relies on an app called Crave. Customers can scan a QR code at their gate or at the restaurant location itself, and then place an order through the app. Rather than putting in their home address, the customer then provides their desired delivery location in the airport, such as their gate number. When the food is ready, it’s loaded into the robot by the restaurant, and then the robot heads off to find the customer. According to Ottonomy, deliveries can be as quick as 10 minutes for orders of retail snacks, or 20 to 25 minutes for those that are freshly prepared by a restaurant.

The robots are being rolled out to several locations for testing, with a pilot program at Rome Airport among various future deployments. The robots were also recently tested at Pittsburgh International Airport, with passengers offered a free drink for helping to trial the system.

Ottonomy obviously isn’t the only company in the robot delivery space. Other notable competitors we’ve seen before include Neubility, LG, and Baemin. It’s interesting to note that many of these companies have converged on similar designs. These robots typically feature the same kinds of sensor packages, drive systems, and overall layouts. The vast majority resemble a cooler on wheels that’s designed to be highly maneuverable and handle regular urban terrain. When your rivals have come up with similar solutions, it can be a sign that you’re on the right track.

Overall, airport food delivery seems like an achievable challenge for robotics startups to solve. In relatively-controlled spaces like airports, the risks are low. With the right packaging choices, the smell from potent curries shouldn’t pollute another customer’s salad. An integrated system for customers to report problems should allow any messy robots to be readily sent back for cleaning in the event of spills, too.

A slow robot with enough sensors not to run over people shouldn’t have too much trouble delivering some food from point A to point B. Perhaps the most likely problem is that the robots will be unable to handle navigating around an airport when passenger volumes are highest. Heavy foot traffic would slow deliveries in these conditions, which would also be when demand is highest. And of course, the time to get back from point B to point A also has to be factored in to every delivery.

At the extremes, if the robots become completely unable to move, they could become frustrating roadblocks for passengers in a hurry. Stress testing a robot delivery system would be ideal, but it’s hard to imagine any artificial test measuring up to the sheer havoc and chaos of Thanksgiving weekend at LAX.

In any case, it seems likely that robot delivery will gradually become more common as the technology matures and kinks are worked out. Get familiar with your robot delivery pals, and be friendly, lest they shake up your soda by using their lidar to find a particularly rough bit of pavement just to spite you!

26 thoughts on “Your Next Airport Meal May Be Delivered By Robot

  1. And they could deliver “internationally”. I had to fly out of Montreal to Europe and a restaurant I liked was only available in the international section, but it wasn’t clear whether it was open, while another that was ok-ish was in the national section. Going to the international one meant no turning back. In the end I ate at the national one, which was meh but the international one turned out to be closed, so good I didn’t go and check there.

    1. Have they shared the costs of purchasing and maintaining the automated carts?

      Suppose you pay a human $15 to push a cheaper manual cart with just the requisite compartments for hot and cold items. If you pay 2-3 humans to deliver orders 16 hours a day, that costs you $240/day, $1,680/week, and $87,360 per year. The actual cost would be higher if you account for employee benefits and other business expenses.

      So whether 2-3 humans are cheaper than 2-3 carts depends on the cost of owning and operating a cart. That cost would include the cost of purchase divided by the number of years the cart is expected to operate, maintenance and storage costs per year, and taxes, among other things. If all of this adds up to less than $87,360/3, or about $30,000/year, than an automated cart could be cheaper than a human.

      The automated carts could also have advantages, e.g., if they break down less often than a human gets sick, or if they need battery recharging less often than a human needs bathroom breaks (probably yes?), or if they can more reliably navigate the airport to deliver orders than a human can. I don’t know if any of these things are true, but they would make a difference beside just the sheer dollars above.

      My guess is that the cost of owning and operating an automated cart will be reasonable compared to a human, so whether this business model takes off will come down to whether they can reliably navigate the airport especially in heavy crowds and whether they can operate without constant technical issues (getting stuck in weird corners requiring a human to assist, or parts breaking and requiring replacement, or the software generally bugging out and requiring engineers to troubleshoot).

      1. It might be cost prohibitive for a restaurant in an airport to afford two or more of these, but if the airport maintained a fleet and each restaurant paid a fee to use the robots, the airport could provide storage for a great many of the little beasts, spread maintenance costs by stocking parts, pass on battery charging and other fees, etc.

        There’s certainly a business model to be made for doing this in a more cost-friendly way. The question is would fliers see having the little buggers running things to gates for customers be a reason to choose one airport over another? Perhaps so if you need that neck pillow for a flight as well as a meal choice for the cross-country or international flight.

    2. Agreed, you cannot beat the price of modern slave worker used by deliveroo/eats : raised on public funds, housed in subsidized housing, working for 10€ an hours. They can solve quite complicated problems, climb stairs, recover from being hussled and move quite fast around any terrain. And best they are disposable, you don’t need to maintain them.

      This self delivering cart is a joke, at this speed it’s next to useless.

    3. @Charles Lamb said: “I like the concept and the research but it seems that having humans push the cart would be much cheaper.”

      Not if the Human is going to shake you down for a superfluous yet hefty tip!

  2. Are there potential security issues around sticking something back in the robot after taking out your food?

    Also 20-25 mins for a meal seems pretty slow; surprised this isn’t much faster given many airport eateries are fast serving.

  3. Is it going to expect a tip ? I can imagine it following someone through the airport because they didn’t tip. What’s worse is having your food delivered to the wrong terminal while your luggage is on its way to Cleveland… 😁

  4. The less human interaction the better. Love the idea.

    Hopefully, they will sedate passengers and transport them like cargo in the future. Thus eliminating the need for seating, catering etc. Prices could be fixed to the weight and size of the cargo. No drunks on board.

    Damm!!! Potential, here…

  5. Seems to me they haven’t solved all the problems. For the final few feet the bot may have to drive/walk/crawl but for the path from food prep to near the person it should be via overhead rail/tram -way. No need for bots to be scurrying among us when it’s possible to avoid.

    1. Reporting on something that is already a thing. Had lunch in Krakow airport, Poland, and it was delivered by robot. It got caught on someone’s luggage and it politely asked them to move it out of the way before it carried on and trundled over to our table with a smile on its tablet face.

    1. A few years ago I heard a relative (with a thick accent) order thick crust pizzas over the phone. He was disappointed when thin crust pizzas were delivered.
      I can understand why the pizza restaurant made the mistake.

  6. When I was flying out of the Philly airport to head out to Supercon this year, there were signs advertising a similar service.

    Was very excited to check it out and almost made an order, until I looked into it a bit closer — because it was still just a trial, a human had to escort the robot the whole way to your gate. So while technically your food would be delivered by the bot, some poor soul had to walk in front of it the whole time telling people to watch out because this lumbering thing was just a few feet behind them. So no human effort was actually being saved, and in fact, you could make the case that it was actually squandering some worker’s time.

    But conceptually, it makes a lot of sense. Surely anyone who’s been in a busy airport knows how valuable a good seat at the gate can be. If I want to go get a drink, it could mean having to stand up when I get back. A bot that could bring me whatever I need so I don’t have to surrender my territory would be fantastic.

  7. lots of potential for shenanigans
    hacked robots could be used for
    drug deliveries
    weapons delivery and removal
    conraband smuggling
    murder hornet assisanations
    in a controlled perimeter it wont be so bad,but in the wild its just begging to be weaponised

  8. They have something similar at Haneda airport in Japan. Although they were robot wheel chairs. The person could use them to get to their gate. Then once there, the chair would return itself. Quite cool I thought.

Leave a Reply to MacAttackCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.