Rui Oliveira

ITSC-Path Planning for Autonomous Bus Driving in Highly Constrained Environments - Results.mp4

Planning for buses

I will present my work "Path Planning for Autonomous Bus Driving in Highly Constrained Environments" at the Intelligent Transportation Systems Conference 2019.

The work deals with planning motions for a autonomous bus driving in urban environments. Unlike passenger vehicles and trucks, buses have a special chassis configuration characterized by large overhangs. To increase the maneuverability of buses, expert drivers will often allow the overhangs to sweep over curbs.

In this work, we mimic the expert bus driver behaviour, by explicitly taking into account the overhangs and allowing them to sweep over curbs.

Optimizing lattice planners

I traveled to China to present my publication "Combining Lattice-Based Planning and Path Optimization in Autonomous Heavy Duty Vehicle Applications" in the Intelligent Vehicles Symposium 2018.

Lattice planners are a commonly used tool to plan motions for autonomous vehicles, however they suffer from sub-optimality and oscillatory behavior. In this work, a new optimization method for reducing these unwanted effects, is proposed, which unlike previous approaches, is implemented in a holistic fashion. The planning and optimization are done in interleaved steps, allowing the planner to benefit from the optimization procedure.

Smoothing shortest paths

I have published an article in the European Control Conference 2018, titled "Trajectory Generation Using Sharpness Continuous Dubins-Like Paths with Applications in Control of Heavy-Duty Vehicles".

This work adapts Dubins paths, a well know concept in robotics, for planning shortest paths between two points, to heavy duty vehicles. Dubins paths are characterized by a discontinuous curvature profile, which requires an infinite steering effort, in order to be followed exactly by a vehicle. Dubins paths are modified so that they comply with the steering actuator limitations, making them easier to perform by autonomous vehicles.

Joining Scania

In 2016, I started at Scania as an Industrial PhD, supervised by Marcello Cirillo, and KTH professors Bo Wahlberg and Jonas Mårtensson.

My research will focus on motion planning algorithms for heavy duty vehicles. Compared to passenger vehicles, trucks and buses face additional challenges when driving on roads, and as such, require additional efforts in order to become autonomous.

Cooperating challenges

I took part in the Grand Cooperative Driving Challenge 2016. The goal of this challenge is to showcase autonomous and connected driving functionality, by performing complex maneuvers, such as platoon merging, intersection crossing and emergency vehicle passages.

Together with my colleagues from the Integrated Transport Research Lab, we developed the required functionalities for this challenge, both for a Scania truck, and for a prototype vehicle, RCV-Research Concept Vehicle.

SML School

In 2015 I developed SML School, a 3-day program, in which middle- and high-schoolers can try out the life of an Engineer. Through practical and hands-on experiments the students are introduced to programming, electronics and automatic control algorithms.

The common topic between all experiments is autonomous and intelligent traffic (see video).

A Royal visit

In the summer of 2015, the Smart Mobility Lab organized showcased the benefits of automated and connected driving to His Majesty The King of Sweden. The demonstration, that can be seen in the video, explains the benefits and challenges of Platooning, a technology which promises increase fuel consumption for vehicles driving close together.

You can read more about the visit here:

iQMatic Demo

I have been involved in iQMatic, a project between Scania, Saab, Autoliv, Linköping University, and KTH Royal Institute of Technology . iQMatic set out the goal to develop a fully autonomous truck, together with a fleet and command center, for mining applications.

At KTH, together with my colleagues João Pedro Alvito and Matteo Vanin, I developed an in-house demonstration of the iQMatic project. The demonstration involves several components, such as autonomous driving functionality and fleet management capabilities (see video).

Backwards towards the goal

In the spring of 2015, I supervised Amro Elhassan's master thesis, which developed an autonomous driving functionality for reversing an articulated vehicle. The work was implemented in a scaled model truck at the Smart Mobility Lab (see video).

Master Thesis

In the spring of 2014, I did my master thesis, which studied motion planning and control algorithms for autonomous vehicles. The algorithms implemented were testes in scaled model trucks at the Smart Mobility Lab (see video).

You can find the thesis here: