We are not afraid to try new things, take calculated risks and find new opportunities

We caught up with the Hexagon team ahead of the Self Driving Track Day series on 10 July. Hexagon is sponsoring the event, they told us all about their thoughts on how the industry can help resolve the shortage in training and recruitment, how they are staying ahead of the competition and the limits to GPS/GNSS performance in today’s most advanced technologies.

Tell us about NovAtel’s relationship with Hexagon and what advantages that has brought to the business.

NovAtel is a part of Hexagon’s Positioning Intelligence division. Hexagon Positioning Intelligence leverages technology and products from its brands NovAtel and Veripos to deliver end-to-end assured positioning solutions. Being a part of the Hexagon family has allowed us to work closer with other divisions within Hexagon to pioneer solutions for emerging markets, specifically autonomous automotive applications.

What advances have taken place in positioning technology in the last 5 years?

Precise point positioning (PPP) is a positioning technique that uses GNSS satellite clock and orbit corrections to model and remove GNSS errors, resulting in a decimetre-level or better position solution. This technique is attractive in many markets because it is globally available and does not require additional hardware or infrastructure for the user.  Although PPP has many advantages, the major technical challenge is the time it takes for the solution accuracy to converge. There has been significant investment in both industry and academia to resolve the technical limitations of convergence time with PPP solutions to the point where highly accurate, instantaneous PPP positioning is a reality today. Hexagon PI has launched TerraStar X technology to address the need for instantaneous convergence in high accuracy PPP solutions. When combined with automotive-grade GNSS receivers available through Hexagon Positioning Intelligence, this technology allows automotive customers to evaluate positioning performance in real-time using data delivered over the cellular network or the L-band frequency.

Sensor Fusion is another topic that is becoming vital to the advancement of positioning technologies. There are many cases where GNSS alone cannot provide an accurate positioning solution so other sensor (LiDAR, cameras, inertial measurement units, etc.) are being used to overcome the limitations of GNSS. Sensor fusion is the concept where these sensors are working together on a platform to contribute to the overall solution. Hexagon PI is already integrating IMUs into our positioning products to deliver GNSS+INS solutions through our SPAN product line. Advancements in the sensors themselves have allowed them to become more practical to integrate, which has led to significant improvements to the availability of a GNSS solution.

GNSS chipset manufacturers are producing automotive-grade multi-frequency GNSS chipsets in anticipation of autonomous driving applications. Dual frequency GNSS with safety and integrity is a critical component to achieving the accuracy and reliability required for autonomy. Hexagon PI is accelerating development in this area with our positioning solutions. Using our positioning engine and combining the GNSS measurements from these chipsets with Inertial Measurement Unit (IMU) data and PPP correction services deliver centimetre-level PPP positioning solutions in real-time.

The availability of GNSS satellites has drastically increased over the last five years to become a truly global system. In addition to GPS, GLONASS, and BeiDou, Galileo became fully operational adding another 18 satellites for navigation. When a receiver utilizes signals from a variety of constellations, redundancy is built into the solution. More satellites mean more signal availability, which allows for reduced signal acquisition time and can reduce the impact of obstructions on the position solution.

How do you feel the industry can help resolve the shortage in training and recruitment?

Hexagon PI supports many university programs related to encouraging youth in their pursuit of GNSS knowledge. We also have a very successful intern program with an 85% post graduate rehire rate. Outside of these official programs we also connect with youth on a regular basis via school visits and social media to promote the knowledge of the industry.

Regarding intermediate to senior level responsibilities we encourage knowledge share within our company through a robust set of programs that encourage creativity and innovation specific to geomatics; such as an internal mentoring program, lunch and learn opportunities and our annual Innovation Week where everyone from the company can take time away from their regular responsibilities to work on innovative ideas that they have.

As Hexagon PI is a global organization with offices all over the world, recruiting is focused on both local and international candidates. Flexible and remote arrangements are common and encouraged to maintain our international presence. Building on flexible immigration programs within Canada, we have been able to attract senior level experienced people who can continue to build our internal knowledge share and elevate the skills of our existing employees.

Does positioning technology augment the vision system in an autonomous vehicle, or the other way around?

The simple answer is both. From our perspective, as a GNSS company, we rely primarily on positioning technology and use external sensors to augment the position solution. Vision systems and GNSS based position solutions are complementary technologies in that when vision systems fail (i.e. weather, poor visibility) the GNSS solution can provide an absolute position, which is crucial to maintaining lane-level accuracy of the autonomous vehicle.

What are the limits to GPS/GNSS performance in today’s most advanced technologies?

One of the major limitations of GNSS is that it does not work everywhere. GNSS positioning relies on the availability of satellite signals and performance is highly dependent on environmental factors. Buildings and trees obstruct GNSS signals, while tunnels and overpasses can completely block the signal. Most GNSS receivers can manage challenging environments through a combination of multipath mitigation algorithms, dead reckoning, and using data from inertial measurement units.

How is your company staying ahead of your competition?

Our range of products and knowledge in GNSS, specifically in safety critical applications, is what differentiates us from the competition. Hexagon PI has products that are uniquely positioned to deliver high precision GNSS solutions through every stage of development in autonomy. Our SPAN products, industry leading GNSS+INS solutions used as truth systems to enable automotive fleet integration, are the foundation for our sensor fusion algorithms.  We recently announced TerraStar X, our technology platform to eliminate the convergence time of high accuracy PPP solutions. We have also worked with STMicroelectronics integrating our positioning engine and correction services on the world’s first automotive-grade multi-frequency GNSS chipsets. Hexagon PI continues to innovate to provide assured positioning anywhere.

Where do ADAS and autonomous vehicles sit in terms of importance among the wide range of applications for your technology?

ADAS and the progression toward autonomous vehicles sit very high in terms of importance for GNSS applications. An absolute position solution that is safe and reliable will be an integral part of an autonomous automotive application. There are still many challenges with ADAS that an accurate position solution can help to solve, especially as we transition to full autonomy.

What are you hoping to achieve as a business this year?

We will continue supporting autonomous driving programs world-wide with our industry-leading SPAN GNSS+INS positioning systems, and to make steady progress towards delivering a functionally safe, ISO-26262 ASIL-B qualified mass production automotive positioning solution based on our software positioning engine and TerraStar-X PPP services in the years ahead.

What would you tell a prospective employee about your company?

At Hexagon PI, we know that the success of our business is a direct result of our highly motivated and collaborative staff. We value our people as much as we value our business. We pride ourselves on providing a stimulating work experience and cultivating teams that encourage learning, so that you can hone your expertise and grow in your career.

We are not afraid to try new things, take calculated risks and find new opportunities. We value performance over procedure, setting measurable goals and working collaboratively to achieve the results we seek.

Some of the perks to working with Hexagon PI include: Expert teams, strong customer focus, flexible work hours, casual dress any day of the week, state-of-the-art work stations and employee led social and environmental committees. Not to mention comprehensive benefit packages, company paid professional development and market competitive salaries.

Come and meet Hexagon in Milton Keynes, UK on 10 July at Self Driving Track Days. Book your tickets here >>

Goal! Lessons learnt from robots playing football

Sander van Dijk, Head of Research at Parkopedia. Delivering a workshop on 10 July on “Getting your Deep Neural Network into a car”

The next in our series of interviews with workshop leaders at Self Driving Track days introduces us to Sander van Dijk. Sander is Head of Research at Parkopedia, and will be delivering a workshop on 10 July on “Getting your Deep Neural Network into a car”

Can you tell us a bit more about your work as Head of Research at Parkopedia?

At Parkopedia our goal is to improve the world by delivering innovative parking solutions. This means capturing and providing any information that is useful to a driver looking for a place to park: locations, both off and on-street, opening hours, restrictions, prices, etc., but also, very importantly, the probability of you actually finding a space in any of those locations. The research team at Parkopedia develops advanced machine learning models to predict this probability accurately worldwide, using a wide range and large volume of input data. Furthermore, we build deep learning-based computer vision methods and other recognition techniques to help automate and ramp up our growth in coverage. Besides that I am part of the team at Parkopedia that works to enable autonomous valet parking.

You have also worked on making humanoid robots play football and are impressively three-time vice champion in the RoboCup world cup. Can you tell us more about how you got into this? Why do has this been such a successful past time for you? What can others do if they want to get involved?

While I was studying Artificial Intelligence at the University of Groningen in The Netherlands, a friend and fellow student mentioned some robotics competition coming up in Bremen, Germany, just over the border, and that we should enter as a team. This was the RoboCup world championship of 2006, and we entered the 3D simulation league, where robots play football autonomously in a physical simulation. We ended up being completely trounced, but we loved the experience and the mix of a friendly, scientific, and competitive community. The next year we participated again, in Atlanta, Georgia, but this time we came more prepared and managed to claim second prize. I have been competing since; I came to the UK in 2009 to start my PhD at the University of Hertfordshire, after meeting my then future supervisor at RoboCup, and became vice world champion again in the simulation league with my new team there, the UH Bold Hearts. A few years later we decided to turn it up a notch, and put together a team of real physical robots to join the Humanoid football league, and in 2014 took the second prize trophy in that league at the world cup in Brazil.

The main aim of the RoboCup is to tackle the very hard problem of having robots and AI operate successfully and reliably in a dynamic world, not just on a discrete and structured chess or go board. This is not only to be able to play football, but will help enable robots and AI in daily life, such as in care, rescue, and autonomous driving. I believe the reason of our success in RoboCup, and of the other amazing teams that kept us from the first place, is robustness. The robots always need to be able to do something, even in the case of failure or unexpected situations, and for the full length of two 10 minute halves. One guideline we follow for this is to only introduce new methods that degrade gracefully: if a new fancy team strategy fails because of broken communication, each robot should still be able to play on its own.

Anybody who would like to get involved in RoboCup can have a look on their website, www.robocup.org, or reach out to our team the Bold Hearts on any of our social media accounts.

Did you learn anything in this hobby that you have applied to your work at Parkopedia?

In RoboCup you often are in a situation having to create/update/fix a complex intelligent system under pressure: a lot of magic happens in the 5 minute breaks between halves. In a dynamic company like Parkopedia it sometimes can feel similar, when combining something unconstrained like research with tight business deadlines and expectations. The aspect of robustness, and incremental, ‘agile’ research helps to keep things under control while making quick progress. The lessons learned from RoboCup will especially apply to our project to create an autonomously parking vehicle. Benefits however flow the other way as well, by introducing industry quality processes to our team mainly made up of students, based on our experience at Parkopedia.

Deep learning is all the hype now, not the least in the world of self-driving cars. Do you think this hype will continue? What do we need to watch out for?

Deep learning is here to stay for a while. It has revolutionised a range of research fields, cutting down the time frames in which even experts in those fields thought such achievements were possible. There are still plenty of fields and problems open to applications of deep learning, and the amount of research output on deep learning is massive. What is very special to this research is the open availability: because the field moves so fast, much of the research is made available openly on platforms such as arXiv.org. This makes it possible for everybody to have access and contribute, not just for the few who are at an institution that can afford to get access through overpriced paywalls.

This is amazingly powerful, but also circumvents the checks of traditional science: although the system of peer review is not without faults, without it it is now very easy to put works out there that look scientific, but may be plagiarized, not possible to reproduce, or just simply bad research. Such works can set false expectations and add negatively to the hype.

Besides this there are still questions around deep learning that haven’t fully been answered: why does it really work so well (it is not uncommon that the original intuitive explanations of some of the smartest authors proved to be wrong later on)? Can we make useful claims about why a deep network makes a decision, especially if something goes wrong? Can we get it to learn useful things from just a small sample of data, like humans do?

In your Self Driving Track Days workshop you will have a look at the different shapes and forms of deep neural networks that are out there that you may want to use on a self-driving car, what is the biggest consideration in choosing the right network design?

The most important, as to any machine learning problem, is to choose a model that fits your problem. This involves seemingly simple decisions like choosing either a classification or a regression network. But a counting problem can for instance be modeled as either, so you have to think about which is more suitable to your use case. An important consideration in that sense is your cost function: what output is good and what output is bad for you? If counts close to the actual value can be good enough, you may choose regression. If any count except the actual value is as bad, you may choose classification. Besides that, it is useful to think about the format of the output to: do you really need the pixel-by-pixel mask of a Mask RCNN, or are bounding boxes sufficient? Especially with limited hardware and real time constraints, it is useful to keep things as simple as possible, and maybe even to reframe your problem to make it possible to run a much cheaper model instead.

What are you most looking forward to about leading this workshop?

I am looking forward to meeting all the people interested in the exciting area of self-driving cars, and deep learning specifically, and to have interesting discussions about the recent developments, opportunities and limitations. I hope the participants will come away with some useful knowledge and ideas for applying these techniques in practice, including myself as there is always more to learn in this fast paced field!

If you want to hear more from Sander, he will be join Self Driving Track Days to deliver a workshop on “Getting your Deep Neural Network into a car” – View the full agenda here >>

Cybersecurity is a mind-set

Madeline Cheah, Senior Cybersecurity Analyst, HORIBA MIRA is delivering workshop on “Understanding Automotive Cybersecurity”

Delivering an in-depth workshop at Self Driving Track Days in July, HORIBA MIRA have taken time out to write an article for us on cybersecurity. Madeline Cheah, Senior Cybersecurity Analyst will be delivering “Understanding Automotive Cybersecurity”


Cybersecurity is really about the absence of behaviour – that is to say – undesirable behaviour.  It is compounded by the fact that it is hard to test for absence, and many of the vulnerabilities – whether that be design choices or implementation errors – are unintended.  How would you test for an absence of unintended behaviour?

There is no such thing as perfect security.  If we use the simple example of car theft – a criminal who really wanted to take a specific car would just pick it up and put it on a truck.  Instead, the aim is to make a system infeasible to attack, in the hope that attackers will go elsewhere.  And if every single system was infeasible to attack, then, in an ideal world, all would be protected.

However, this is not an ideal world, and what we have instead is a footrace between attacker and defender.  This is exacerbated by the attacker-defender imbalance, whereby the attacker only has to find one vulnerability to exploit, but a defender has to protect as much of the system as possible.  There is always a balance to be struck between what’s usable, what’s cost-effective, what’s reactive and what can be proactive.

Automotive Cybersecurity

There are several drivers in the automotive industry that have led to challenging issues in the cybersecurity arena.

Firstly, there is increased connectivity, both inside and outside the vehicle.  Secondly, there is increased complexity, with the advent of many systems that have been introduced for reasons of safety, security or marketability. Finally, there is convergence of technologies, between those that were designed for vehicles (for example advanced driver assistance systems), and those that were integrated into vehicles (wireless Internet connectivity).

There are no “strong” or “weak” parts of a vehicle; even what might seem like a trivial attack can be chained with other attacks to make the end impact potentially devastating.  Furthermore, with every feature that makes something more safe or convenient, there is potentially an equal amount of convenience for an attacker if sufficient defences are not in place.

Instead we talk about hardening the system, where we close as many security holes as feasibly possible.  There are some fundamental principles we can follow as a guide. We can use defence in depth, where there are multiple layers of security, such that holes in one layer are covered by another layer.  We can use the principle of least privilege, where by default, nothing is allowed, with the necessary functionality enabled one by one.  We can ensure that security is in a system by design, rather than being retro-fitted, such that the holes don’t appear in the first place.

System Boundaries

Currently, we can realistically draw a system boundary around the vehicle for the purposes of testing or analysis.  However, the horizon is full of vehicles that are connected to each other, to the cloud, to infrastructure, to peripheral devices and wearables. The vehicle will not just be the end target, but the means to an end, whether that be for a backdoor into infrastructure, ram-raiding, terrorism, privacy violation, financial crime and all other criminal activities that now take place through more conventional computing methods.

Cybersecurity in many ways is a mind-set.  Not everyone has to be aware of every technique or attack, but knowing when and where to involve a security engineer is crucial. Even if the actual product itself is purely mechanical, sooner or later, it will be attached to or integrated with an electronics system.  Presumably, the use of computing and IT is used for designing the product, for protecting the intellectual property attached to a product and for data analytics. Away from automotive technology, cybersecurity awareness can take many forms in many disciplines.  In design disciplines, we look at developing security that lay users can use to protect themselves. With psychology and linguistics, it could allow us to distinguish various types of threat actors.  Understanding of the law could help with questions to do product liability.


Here at HORIBA MIRA, we aim to be a trusted partner to manufacturers. That means working collaboratively with the project team every step of the way. Our cybersecurity related projects cover consultancy, concept development (security by design) and independent security assessment, whether at component, vehicle or lifecycle level.

We are also active in the research sphere, whether that be through collaborative projects funded by InnovateUK (such as 5Stars), through applied research internally, or through embedded PhD programmes in the business.  HORIBA MIRA is also heavily involved with the development of international standards in the field, with our experts representing the UK in the development of ISO/SAE AW 21434 (Road Vehicles – Cybersecurity Engineering) and ISO26262 (Road Vehicles – Functional Safety).

Join HORIBA MIRA at Self Driving Track Days in Milton Keynes this July. Book your tickets to attend >>

New Autonomous Vehicle Zone at the London Motor Show 2018!

Hi everyone!  Just me checking in with some fantastic news that I felt obliged to write about. 

Along with AutoSens, our sister event series in the international vehicle perception sector, we are taking over a major part of The Confused.com London Motor Show 2018 with a brand new Autonomous Vehicle Zone.

It’s remarkable to think that we’ve come so far in only 16 months since we launched, and that we’ll leapfrog to join the likes of the largest auto events in the world to tell the story of the future of transport.

I have started chatting to people about this announcement over the last couple of weeks, and it’s clear that 2018 is going to be a fantastic year for the sector, particularly in the UK – and that means we are expanding too, as our in-house team approaches double figures.

New courses are coming (both accredited and online), new technology is coming, recruiters are getting busier and investment is accelerating across Europe, so it’s a great time to be thrust into the public’s imagination to help changing hearts and minds, talking about the challenges and opportunities ahead, and sharing the journey.

This announcement will be seen more widely by the public, the mainstream and automotive media over the coming days, and of course we are keen to flesh out the announcement with more details of the features, exhibitors and exciting discussions that we’ll be able to bring to you in May 2018.

Finally, a treat for those of you that are avid fans.. ticketing for Self Driving Track Days will be live soon, and we’ll be giving away one place at the July event in Milton Keynes before Christmas to one lucky person that joins the mailing list in the first two weeks of December 🙂

Have a great holiday season!

Alex Lawrence-Berkeley [LinkedIn]

Co-Founder of Self Driving Track Days and Head of New Projects at Sense Media Group.

Austria 2017 review

ÖAMTC Fahrzentrum Teesdorf played host to the final Self Driving Track Days event of our inaugural 2016-2017 series.

In the region of Baden, Austria, only a short journey from Vienna, around 90 people attended this exciting one-day event, experiencing demonstrations in vehicles provided by TU Graz and Virtual Vehicle Research Center, both based in the country, and learning from engineers and technology developers working on autonomous vehicle technology.

Support from industry

Workshop sessions were hosted by 3-hour sessions were delivered by specialists from TU GrazCodeplay SoftwareAIMotiveHoriba Mira, with further contributions from NovAtelQuantum, and a joint demonstration from NXP and Intempora, with valued support from DataSpeed in the exhibition, demos and panel Q&A session chaired by co-founder Alex Lawrence-Berkeley at the end of the day.

International event

Attendees from Austria, as well as travellers from Poland, Germany and the UK, attended talks by speakers from countries including the US, UK, France and Hungary at a unique and buzzing event which received very positive feedback from sponsors, speakers and attendees alike.

What attendees said

  • “Organization is perfect… price to performance ratio is outstanding” Tomasz Bialek, FEV Polska
  • “At Self Driving Track Days Austria I got to meet several major companies in the field of autonomous driving. I had a blast, especially while catching a ride with a self-driving car.” Kaspar Sakmann, T-Mobile Austria.
  • “It was meaningful to know and experience the state-of-the-art self driving technology. First of all it was good chance to take a test run on autonomous vehicle. In addition to this event was very well organized.” Chansong Jeong, FH Technikum Wien.
    Arpad Takacs, Outreach Scientist at AI Motive, delivering the “AI in the ecosystem of self driving cars” workshop
  • “Good talks, interesting in-car demos and networking opportunities.” Sebastian Busch, Elektrobit Austria GmbH
  • “A super one day event with a wide variety of information.” Christian Aschauer, Universität für Bodenkultur,Institut für Landtechnik.

Outreach and visibility in the community

It was evident from where the attendees had heard about the event that Self Driving Track Days is reaching groups, companies and organisations that typically don’t access international events, including various meetup groups, small companies and researchers in smaller companies and academic organisations.

Arno Eichberger, or TU Graz, being interviewed by Austrian TV

Media coverage

Broadcast, online and print media were in attendance, with features online, on national TV and on radio by the Austrian national broadcaster, ORF, as well as online features in industry magazines Fleet & Economy, Car & Economy, international online coverage on IEEE Spectrum’s Cars That Think. See more…

Support through the year

It’s important for us, as well as the hundreds of people that have engaged with demonstrations, workshops and events we have run in this first year to take a moment and thank all of the companies that have made this possible:

Firstly out partners and sponsors, AutonomouStuff, NovAtel, Quantum Corporation, Intempora, NXP, LeddarTech – without whom Self Driving Track Days would not have happened at all.

Our friends in the industry that have worked so hard to provide demonstrations on-track, including DataSpeed Inc, Anthony Best Dynamics, Virtual Vehicle Research Center and TU Graz

Many other people that have helped along the way, such as journalists, numerous guest speakers at our meetup and track events, marketing teams, venue crew, contractors and colleagues, insurers, camera crew and especially all the meetup group organisers that have helped promote these unique events to their communities of interested people – above all others, as volunteers, they have made the biggest difference – please support them!

What’s next

A note from Alex:

“It’s been a crazy year, a very steep learning curve and utterly exhausting.  I am grateful to my colleagues in Sense Media, especially Rob Stead, for supporting this challenging project and I am looking forward to seeing it grow and mature in such a short space of time.  

Self Driving Track Days will return with more events in 2018, but if you can’t wait for your autonomous technology fix, please sign up to the mailing list for updates and news about future events, and check out AutoSens this September in Brussels, which is the much bigger industry event we also organise. See you next year!”