From the Editors

Within the AV ecosystem, passenger vehicles generally get the most attention. What is often overlooked is the significant potential of automated vehicles for local delivery, long distance freight, and pilotless electric planes for short- to medium-distances. There are similarities in the underlying technologies in all of these. They all need various sensors, artificial intelligence, position, navigation, and detect-and-avoid strategies to avoid other vehicles.

This issue of AV update includes items on all of the above.

---

Canadian AV News
Canadian Standards Association operating as CSA Group with support from Transport Canada’s program to Advance Connectivity and Automation in the Transportation System (ACATS) is engaging stakeholders to develop standards and guidelines, contribute to international standards development, and develop a forward-looking standardization roadmap. These activities will help support the safe, secure and efficient deployment of CAVs across Canada. To guide this process, CSA Group has convened a Connected and Automated Vehicle Advisory Council (CAVAC) which will review the codes and standards landscape and formulate strategy for developing best practices, guidelines, and standards for CAVs. CAVCOE's Barrie Kirk is a member of CAVAC and the first meeting was held in May. This should be a very interesting and important project.

---

Transport Canada organized and hosted a Forum on Low-Speed Automated Shuttles in June 2019. The presenters and attendees were a who's who of stakeholders involved in this space, including shuttle developers / manufacturers, service providers / operators, representatives from all three levels of government, and a representative from CNIB. CAVCOE's Barrie Kirk and Marie-France Laurin were also there. The presentations and discussions were excellent and described previous pilots, current activities and future trends. For us, the most useful takeaways were the comments from multiple speakers on the best practices for planning and deploying pilot projects.

---

It is often predicted that AVs will make traffic congestion worse once they go mainstream. Some  experts at the Victoria Transport Policy Institute in B.C., the University of Waterloo and others appear to agree with this point of view. Some even go so far as to say that communicating AVs could conspire together to cause congestion to avoid paying for parking!  More information is at this link.

---

On June 12, 2019, the Globe & Mail published an article titled the shift to a self-driving, electric future may not be quite so automatic. The article cites some transportation modelling work conducted at the University of Waterloo suggesting the optimistic forecasts for the widespread acceptance of electric and self-driving cars by the public may not materialize. The issue boils down to people’s habits about how they use the transportation system and their resistance to radical change.  The article can be viewed at this link. 

---

Caliber Data Labs is a Silicon Valley startup with an engineering office in Canada (Vancouver).  This company develops the picks and shovels for the AV industry. These include workflow and infrastructure tools for data collection, enrichment, annotation, and model deployment by utilizing the latest best practices in production-ready AI and machine learning. More information is at the company website at this link. Also check a recent article by Caliber titled What makes vehicles "smart"? at this link or at this link.

---

As we mentioned above, automation is not limited to ground vehicles only.  CN Rail has adopted automated technologies that can save it up to $100 million per year in its operations. A typical rail-car inspection takes about two minutes to completer by a mechanic shining a flashlight on a car’s undercarriage to spot defects in need of repair. New automated inspection portal using high resolution cameras and artificial intelligence can evaluate 120 cars in the same time it takes a worker to check a single car. CN has already installed four of these automated systems in its network in Canada with more planned in the near future.  CN is also using an automated system using LiDAR for inspecting the tracks and ties at normal operating speeds. More information at this link. 

---

International AV News
Postmates is a major food and goods delivery company based in California. It has been in operation since 2011 and has presence in nearly 3,000 U.S. cities. It has now designed its own delivery robot called ‘Serve’ to complement its large number of conventional vehicles with human drivers. Serve can carry 50 lbs (22.6 Kg) and go 30 miles (48.2 Km) on a single charge.  Customers interact with Serve using a touchscreen and cameras. More information is at this link.

---

In a similar move, Domino’s Pizza announced a partnership with delivery robot specialist – Nuro to use Nuro’s R1 delivery vehicles to deliver pizza in selected areas of Houston later this year.  Nuro has also been using the R1 to deliver groceries to residents of Scottsdale, Arizona, and more recently, Houston. More information is at this link. 

---

The United States Postal Service (USPS) is running a pilot project with self-driving truck startup TuSimple to evaluate delivering mail and packages using self-driving trucks. The pilot project operates between Phoenix and Dallas covering a distance of over 2,100 miles (3,380 km) or around 45 hours of driving. More information is at this link. 

---

Many surveys have been conducted to gauge people’s attitudes towards autonomous vehicles.  One of the latest was released on June 4, 2019 by the U.S. engineering firm HNTB.  Over 1,000 Americans were surveyed in this study.  Respondents had the most concern over liability in collisions, with 56% identifying the issue as a point of concern. Cybersecurity, or the possible hacking of the vehicles, came in at a close second at 53%. People also expressed concern about the initial costs of updating infrastructure for the vehicles, equal access to them and continued funding. More information is at this link.   The survey results can be viewed at HNTB’s website at this link.

---

Gatik is another AV startup focused on short-haul logistics defined as a range of 200 miles (322 Km).  Gatik’s AV technology is designed to fill the gap between automated 18-wheeler (Class 8) trucks suited for highways and small delivery robots designed to operate on sidewalks.  Gatik has partnered with Walmart to test its automated vehicles and technologies. Gatik claims that goods can be transported up to 50% less expensively in city environments between locations.  More information is at this link.

---

Ride-hailing company Lyft has reported that it has carried about 50,000 passengers in its autonomous vehicle fleet in Las Vegas in the first six months of 2019. This makes Lyft the largest commercial self-driving car operator in the U.S.. Lyft has a fleet of 30 BMW 540i in Las Vegas which are fitted with sensors, controls and software by technology company Aptiv. Lyft also states that passengers gave their automated ride an average rating of 4.97 out of 5.  More information is at this link.

---

In addition to ground vehicles, automated vehicles also include air and marine vehicles. In partnership with XPRIZE Foundation, Shell Oil had sponsored a US$7 million for Advancements in Autonomous Ocean Exploration.   On May 31, 2019, XPRIZE announced that the US$4 million first prize was won by GEBCO-NF (International). The winning team integrated existing technologies and its ocean-mapping expertise to develop a robust and low-cost unmanned surface vessel (SeaKIT) that allows for rapid seabed visualization. The ultimate goal is to complete comprehensive mapping of the ocean floor by 2030.   More information is at this link.

---

Using machine learning techniques, MIT researchers have simplified the way an AV can navigate unfamiliar roads.  Traditional navigation systems process data from sensors through multiple systems to achieve localization, mapping, object detection, motion planning, and steering control. The MIT technique trains the machine learning algorithm on how a human driver steers the vehicle on a typical roadway. Once trained, the vehicle can then navigate unfamiliar roads based on what it has learned from the human driver. More details are at this link.

---

Transportation futurists -- and science fiction writers -- have long dreamed of ‘flying cars’ as an alternative to automobiles. Many firms are attempting to make this dream come true. Now, ITS America has announced plans to run demonstrations of electric vertical takeoff and landing vehicles (eVTOLs) at the ITS World Congress in Los Angeles in 2020.  ITS America predicts that such vehicles could be in public use by the time the Olympic Games arrives in LA in 2028. More information is at this link. 

---

Aurora Innovation, a leading AV company headed by AV industry veteran Chris Urmson continues to be in the headlines.  Fiat Chrysler Automobiles (FCA) has now formed a partnership with Aurora to incorporate Aurora’s Level 4 self-driving technology into FCA’s commercial vehicles. Aurora also has partnerships with VW and Hyundai. In another development, Aurora acquired leading LiDAR manufacturer Blackmore Sensors for an undisclosed sum. More information on the FCA partnership is at this link and on Blackmore Sensors here. 

---

Uber has had a long partnership with Volvo for its AV program.  Uber has now announced a third-generation version of its self-driving car, developed in partnership with Volvo. The new XC90 SUV will be built with Uber’s self-driving technology at the factory level, instead of needing to be retrofitted like previous versions of the car.  To make the new vehicles safer, they will have multiple redundant backup systems for critical systems such as steering, braking, and battery power. More information is at this link.

---