Next time you see one of these little autonomous cars in your neighbourhood, stop and say hi, you never know, it might offer to help with your groceries, or it might need your help to reconnect a lost child with his or her parents. In fact, by 2035, more than 12 million fully autonomous vehicles are expected to be on our roads.* But, who says they all have to be big?
In June, Arm and Hackster.io asked the developer community to create an autonomous robot that could roll, swim or fly. Four months later, we have a number of successful autonomous robot projects submitted. We have been blown away by the effort and the commitment shown by the different participants and we are super excited to announce the winners to our challenge in this blog.
The challenge was broad; an autonomous robot that can help in addressing a real-world scenario. We wanted developers to push the limits of what low-cost, open-source hardware and deep learning can do for humanity, addressing social and environmental issues, as well as exploring heterogenous Arm based solutions.
To jump-start the challenge, we asked everyone to submit their best idea. A month later we judged 203 submitted ideas and gave the top 50 proposals a DonkeyCar, an open source DIY self-driving platform, to start developing their solutions. Out of these, the top five projects also received a Avnet Ultra96 FPGA board. Some contestants also used other hardware platforms developed by our Innovators, including OpenMV, JeVois and the Pixhawk4 flight controller.
We gave the community the chance to develop an autonomous robot based on their platform of choice and the results have been astonishing! Applications include underwater robots that monitor water quality and drones for mapping pollution in real time.
Many of the submissions built, use the DonkeyCar. The platform, in most cases, was enhanced by optimizing and adding features to tackle the needs of specific applications. Some of these applications are delivery systems, rovers that asses diseases in plants and robots that can be used to quickly find survivors in disaster situations.
The level of the prototypes has been above our expectations. The technical depth and the extent to which the various participants documented their projects has been fantastic. We asked five judges to evaluate these projects, and after many sleepless nights, we have reached our difficult verdict.
The winners have been placed into four categories (listed below in no particular order). Make sure you click on the project links to see the story behind each robot and the guide to building one yourself!
We have also rewarded five participants with a GoPro for the effort they have put in to documenting their progress and sharing it on social media throughout the challenge!
If you haven’t had a chance to build a rover or a drone, make sure you check out these guides and give it a try. You never know, the next rover that will help future natural disasters, could be the one developed by you!
To hear about future contests, sign up to hear the latest from the Arm Innovator Program.
Sign up
*Statistic from the Boston Consulting Group, Inc. (BCG)