Interviews, Smart Mobility

Interview with FEV North America on functional, system & operational safety for level 4 & 5 automated vehicles

OSS.5 USA is America’s first platform bringing together all stakeholders who play an active role in the achievement of high-level functional, system & operational safety for fully automated driving.

In the run-up to the event, we.CONECT spoke with David. A. LaRue – Technical Fellow – Autonomy and Functional Safety at FEV North America Inc. about new technical innovations, latest updates to standards, and pressing challenges regarding operational & functional safe systems in Level 4 & 5 vehicles.

we.CONECT: What are your main responsibilities in your current role?

David. A. LaRue: I serve as a technical expert regarding safety in the development and integration primarily of autonomy or HV electrification applications. This includes working in the development of sensors and systems up through performance and evaluation at the vehicle level.

we.CONECT: What technical challenge fascinates you most about autonomous driving?

David. A. LaRue: First; development of a reasonable sensor suite that can reliably support the level of situational awareness required in support of L4/L5 autonomy.  Second: doing so in a way that always maintains the vehicle in a safe state despite the presence of a fault or an issue which is limiting the system’s ability to provide the needed situational information required.

we.CONECT: Where do you see the biggest opportunities in the application of AI, machine-, deep- & reinforcement learning in the development of fully autonomous vehicles?

David. A. LaRue: AI/Machine Learning will allow for an opportunity to understand the almost infinite environmental variables that are present in the real world.  This will provide information to close the gap of those things we don’t know we don’t know as well as to finds way to better optimize the overall system to improve performance and overhead.

we.CONECT: How effective are virtual testing, simulation and modelling in Level 4&5 today? What should be improved?

David. A. LaRue: Virtual testing of any type is only reflective of those environments and situations to which they have been programmed.  Virtual capability will continue to improve as we grow in our understanding of the fluid, real-life environment and situation experienced in support of full autonomy.

we.CONECT: How does a completed Safety Case could look like for an L4+5 Autonomous Vehicle? 

David. A. LaRue: Specific to L4/L5, until such times as they are required, standards for implementation and validation of fully autonomous systems, the needed Safety Case will have to be a complete list of not only what is required for proper implementation but will also include considerable detail around the specific environment to which it can and cannot support the autonomous feature.  This will have to include many aspects of weather, road condition, location, traffic density, time of day as well as a number of other variables that may affect the performance of the system.  This means not only in monitoring and mitigation of a fault but the ability to determine the systems’ performance even where there is no fault present.  The sensor suite can see a reduction in performance without a fault present due to weather or other variables, the level of which may be dependent upon the specific sensor technology and the wavelength in which it operates.  As an example, an IR (infrared light) video sensor can be helpful at reducing ghost targets (false positives) as it operates in a very narrow wavelength (~800 nm)., however its performance can be greatly reduced in the presence of fog.  A challenge will be to reduce and account for those influential variable we don’t know we don’t know.

we.CONECT: What have been the highlights of your work with AI/machine learning in the development of autonomous vehicles so far in your career? 

David. A. LaRue: My greatest highlight in this area is the shear value and complexity of what has to be collected, how to ascertain the importance/priority of that collected and lastly how to confidently make use of this vast data set in the development and validation of these functions/features.  I find the complexity and non-linear control challenges most rewarding.

we.CONECT: How is your company developing its AI/machine learning/deep learning capabilities? What are the challenges?   What are the challenges?

David. A. LaRue: FEV has a global ADAS/AD team that has developed a Level 3+ vehicle as an R&D platform. One of the areas of interest that the FEV team worked on was developing AI/machine learning/deep learning algorithms to detect and classify traffic signs and moving objects such as vehicles and pedestrians as well as predicting the trajectory of the moving objects.  Some of the challenges that the ADAS/AD team is facing is getting and determining the mot influential datasets with the right scenarios, computing power, and validating the developed algorithms.  This is supports our edge case research.

we.CONECT: What would you change in the AV industry to speed up and improve autonomous R&D here? 

David. A. LaRue: We have many professional, creative minds across the globe looking at AV from many different perspectives.  However much of this work is built off of antiquated vehicle control architectures.  Our current ECU centric architecture is resulting in added cost, excess capability, greatly increased power consumption and unnecessary system complexity.  A sensor centric system would reduce complexity, power consumption and cost while improving system response/latency, sensor fusion  as well as improve overall computer resources optimization.  This, as well as the use of DCU (Domain Controller Unit), CDD (Complex Device Drivers) DDS (Digital Data Service) and safety wrappers will help to reduce the overall computational burden while improving the safety coverage of the system.  There are other strategies that can also leverage information outside the vehicle environment, not only from infrastructure but from other vehicles, especially in high traffic density situations. 

we.CONECT: What expectations do you have towards the OSS.5 USA 2021?

David. A. LaRue: OSS has attracted many of the industry’s top technical experts and decision makers in all areas of autonomy.  For me, having a direct opportunity to meet with and discuss the challenges before us with these experts is very insightful and rewarding, it also provides an opportunity to understand these challenges from vastly different perspectives. 

we.CONECT: Thank you for the interview! To all our readers, if you are interested and want to join us together with FEV and want to learn, engage and discuss automotive tech innovation in real-time with thought leaders across the globe, feel free to save your seat: 

Previous ArticleNext Article