Hi everyone! It’s Sahil, Luke and Nathan with another founders update on our progress over the last few months. The team has produced some fantastic results this quarter, specifically in regards to hardware development.
We have upgraded the sensor box, which houses sensors, power electronics and the computer system our platform will need to perceive, guide and control a mobile robot to complete its desired objectives. Due to this upgrade, we can now collect large amounts of data in real-time, which includes the vehicle state and surrounding environment.
We will not only be testing in stationary environments, but also dynamic environments, with our colleagues moving within the workspace of the vehicle. The ability to operate within such dynamic environments is extremely important, as this will increase not only the safety and reliability of the product but allows for synergy to occur between the vehicle and individuals working in the surrounding environment.
A remote-control system will be used to do initial test runs which we will use for data collection through all our sensors, allowing us to then run this data through our algorithm simulations before we implement them in the real world. We have wired up emergency bump sensors and a new LED light system, so we can easily identify faults and behavioural issues with our system such as triggered sensors.
Nathan (CTO) has some initial test results on how accurately his navigation algorithm can track moving objects in our vehicles environment. Based on our testing we have found that our algorithm can compensate for unpredictable obstacle movements much better than the current best methods which is a MASSIVE breakthrough.
The graph shows this comparison for the normalised error of both position and extent (error between the estimated size and orientation of the object compared with the true size and object shape) of an object that makes sudden movements (shown at scan# 40, 80 and 110). As can be seen by the graph, our algorithms normalised error [in Orange] is significantly less than the current best method [in blue] and during the sudden movements it quickly corrects its estimate.
Over the next quarter we shall continue testing our smart navigation algorithm, performing further tests in both static and dynamic environments. Current results have been extremely promising, and we are excited to showcase a prototype once it has been completed.
Jake (Communications & Web Dev) has given our website a much-needed revamp. Use the navigation bar at the top of the page to learn more about Elite Robotics and Autonomous Vehicle Technology.
Sahil (CEO) has been busy with fundraising and the partnerships side of the business. In between this he has also had the opportunity to present at many events this last quarter including talks given to UON STAR4000 (Entrepreneurship and Innovation) & Centre of Organic Electronics students, a livestream talking about our startup experience during and post Icon Accelerator, as well as being a speaker at the last Lunaticks Live event (check the video out here) where he spoke about Intelligent Transport and the future of autonomous vehicles. In addition to this, Sahil also gave a short talk about his experience during the China Start Program. Don’t worry if you missed out on this because he will be giving another presentation during Newcastle’s China week at the end of this month. Register here