Nobody likes traffic, and while someday soon we might be chartering flying taxis to zip over it, cities like Amsterdam have another alternative – an intricate network of canals. The RoBoat project is planning to get autonomous boats on those waters to ferry passengers and cargo, join together to form temporary floating structures, and monitor the environment. Now the design has been tweaked to be make them smarter, more agile and easier to manufacture.

The RoBoat project is a collaboration between MIT and the Amsterdam Institute for Advanced Metropolitan Solutions. The goal is to ease congestion on busy streets by passing some of the traffic load to the underused canals, with the autonomous boats acting like taxis, delivery vans, garbage trucks or whatever else they might need to fill in for. And carrying out some of those city services at night could help clear the roads even more.

"Imagine shifting some of infrastructure services that usually take place during the day on the road – deliveries, garbage management, waste management – to the middle of the night, on the water, using a fleet of autonomous boats," says Daniela Rus, co-author of a new paper describing RoBoat's progress.

To achieve this, the project team developed smart little boats kitted out with a range of sensors and electronics, and gave them the ability to join together to build temporary floating structures like bridges, jetties, or even large platforms for concerts and markets. The RoBoat team trialled a prototype of these boats back in 2016, sending it along a preprogrammed path through Amsterdam's canals.

Now, the team has improved the design. First and foremost, manufacturing was made easier by 3D printing the hull. The job took 60 hours using a commercial printer, which produced 16 pieces that were then joined together and sealed with several layers of fiberglass to form a hull measuring 4 m by 2 m (13 ft x 6.6 ft).

Each of the boats carry a power supply, Wi-Fi antenna, GPS, an inertial measurement unit (IMU), a minicomputer and a microcontroller. The GPS modules can pinpoint its location within centimeters outdoors, while indoor tracking is achieved thanks to an ultrasound beacon system. The IMU keeps it from tipping over and measures which direction it's facing.

To make the boats more nimble, the researchers placed the thrusters in the center of each of the four sides. That lets the boat move laterally as easily as it does forwards and backwards. The other key innovation, the team says, was the algorithm driving the vehicle, which was designed to take into account factors like centrifugal forces, drag and changing mass as the boat accelerates or slows down.

"The controller considers the boat dynamics, current state of the boat, thrust constraints, and reference position for the coming several seconds, to optimize how the boat drives on the path," says Wei Wang, first author of the study. "We can then find optimal force for the thrusters that can take the boat back to the path and minimize errors."

To make sure the new control algorithms were working, the researchers ran tests with a miniature prototype following preset paths in a swimming pool and a river. Over 10 test runs, its tracking errors were smaller than would be expected using other algorithms.

Future work will involve finding ways to account for waves, currents and the added mass of passengers and cargo.

The research paper is being presented at the IEEE International Conference on Robotics and Automation this week, and the team demonstrates the RoBoat tests in the video below.

Source: MIT

View gallery - 2 images