Autonomous Mini-Shuttles Why Autonomy? CALSTART Webinar April 18, 2017 Michael Ippoliti, CALSTART

Autonomous Mini-Shuttles Why Autonomy? CALSTART Webinar April 18, 2017 Michael Ippoliti, CALSTART

How Autonomy is Achieved Common technology is ready; development process will take time Cars, Trucks, Buses, Drones (UAVs), AGVs (material handling) Autonomous Vehicles versus Connected Vehicles Not the same - autonomy does not require connectivity, but greatly benefits Components of Autonomous Vehicle Systems See, Think, Act Sensors: cameras/imaging, radar, lidar, sonar (+ connected data, ultraprecise maps) Sensor Fusion: Powerful Processors (e.g. NVIDIA DrivePX2) Control Decisions: Software, source of brand differentiation Actuation: Manipulating the vehicle; advantage to EVs

Levels of Autonomy (SAE J3016) Stages have been defined; development may not go in order These shuttles are LEVEL 4-5 vehicles no need for driver, can not be driven

Why Shared? The Role of Transit Transit must play a large role, if we are to achieve the promised benefits from autonomous vehicles Shared-use mobility along with automation raises new possibilities for sustainable transportation systems that could meet goals for safety, traffic movement, energy efficiency, and emissions that previously appeared to be unattainable. TRB Transformational Technologies in Transportation May2016 However, the impact of autonomy on traffic and congestion is more complex than just making driving itself more efficient. Though automatic driving should increase capacity, we have known for a long time that increased capacity induces more demand - more capacity means more traffic. If you reduce congestion, then more people will drive, either taking new trips or switching from public transport, and congestion might rise back to where you started Benedict Evans blog post 3/29/17 Heaven and Hell Scenarios VMT down, congestion down, emissions down significantly, commuting easier VMT up, 0-occupancy trips, congestion up, emissions slightly better, longer commutes Ride-Shared Car vs Mini-Bus vs Full-Size Bus Future customer demand?

Sharing = Transit Transit is facing competition from new mobility you are the incumbent and can (MUST) leverage the new technologies Sharing needs to become a more mainstream behavior To see VMT stay on that downward trend for millennials, and perhaps creep down for other generations, fleet-based autonomous vehicles will be key. We could be in a lot of trouble congestion wise if people want to own their autonomous vehicles. Most of the studies I ve seen have shown there will be an increase in VMT if we stay the course. It gets a bit different with a fleet model. Fleets could keep us from adding lanes and use thinner lanes and could allow us to add things like sidewalks and bike lanes. 2016 Strategy Analytics report The public sector and transit have to collaborate in the technology development Transit needs to be a big part of the thinking for becoming an AV society, and that too falls under the public sector s purview. When you re going to have more bandwidth of cloud data running into cars than what exists in many homes, how can transit possibly compete with the user experience of autonomous vehicles? If AVs can come to market following these models, as car-sharing and transit partners work to shift cultural preferences about ownership, it s certainly possible that the new technologies will begin to create more efficiently flowing highways and streets. Mobility Lab, June 2016

Safety Safety data for transit is not as well studied as light duty and trucks reducing accidents is always a goal. These systems WILL reduce accidents. In light duty, CAVs could save 30,000 lives a year in the U.S. at a cost savings of 30 percent to 60 percent per mile. Avery Ash, INRIX, Empire State of Mobility Conference 4/12/17) CAVs promise the elimination of driver-related factors in accidents 360 degree view, with no distractions In 2015, at least one driver-related factor was recorded for 33 percent of the large truck drivers in fatal crashes, compared to 57 percent of the passenger vehicle drivers in fatal crashes. (Bus?) Speeding, Distraction/Inattention, Impairment FMCSA Large Truck and Bus Crash Facts 2015

Safety SAFETY is the #1 reason for these technologies starting with ADAS, leading to full automation. This technology will reduce accidents. On average, from 2005 to 2015, intercity buses accounted for 13 percent, and school buses and transit buses accounted for 41 percent and 33 percent, respectively, of all buses involved in fatal crashes. Over the year 2014 to 2015: The number of buses involved in fatal crashes increased from 236 to 261, an increase of 11 percent, and the bus involvement rate in fatal crashes increased by 9 percent. Bus VMT increased by 1.4 percent FMCSA Large Truck and Bus Crash Facts 2015 Increasing accident frequency? Growing need for driver assistance and task automation?

Why EVs (zero-emission)? EVs are better for EMISSIONS, ENGINEERING, and RELIABILITY reasons Boston Consulting Group says the convergence of three trends ride sharing (services such as Uber and Lyft), autonomous driving, and vehicle electrification create a far more compelling economic case than any of these forces alone. Electrification will be the lever for this change because of its positive interactions with the other factors. In the case of autonomous vehicles, electrification is beneficial because the extensive volume of sensors and processors required to operate effectively requires substantial amounts of electric power, which is best provided by an electric vehicle https://www.bcg.com/d/press/10april2017futureautonomouselectricvehicles151076

Automated Technology links with EV Technology Because (6 reasons): 1) ADAS/Automated and EV technologies are developing together The technologies will mature at the same time, and light duty will be leading the shift to electric, bringing with it pressure on fueling and operations, pushing trucks/buses to adopt electric drive. For example, wireless fast charging will eliminate plugging in, and work beautifully with autonomous operation in depots and yards. 2) Easier Engineering as heavy vehicles get more Advanced Driver Assistance Systems (ADAS), and more X-by-Wire, it becomes easier to engineer from an EV base than to re-design a legacy internal combustion engine design (e.g. 24v? 48v? etc.). 3) Reliability/Ruggedness The number one performance parameter for commercial vehicles. Even though hard data is scarce right now, the basic logic is pretty clear. Solid-state components, no oil, far fewer moving parts, simply must result in greater reliability and maintainability. Power and torque will be superior to ICEs. Once past prototype stage, EVs will easily beat ICEs for uptime. 4) Regulatory/Emissions in urban areas especially, it is likely there will be regulated ZE zones in the near future. Even without ZE zones, there will be tighter and tighter ICE emissions restrictions, meaning the cost and complexity of ICE emissions control systems will skyrocket. EVs make zero emissions with no additional un-reliable after-treatment equipment. 5) Regulatory/Mandates Norway and Germany have already hinted at banning ICE cars. It is entirely reasonable for governments to begin seeing ICEs as irresponsible when fully Zero-Emission options are cost competitive. Trucks and buses will be even bigger targets. 6) Cost People like Tony Seba (http://tonyseba.com/portfolio-item/clean-disruption-of-energytransportation/) and now even McKinsey and Bloomberg New Energy Finance (https://about.bnef.com/press-releases/electric-vehicles-to-be-35-of-global-new-car-sales-by- 2040/) have been clear battery prices will continue to drop, and somewhere about 2025, it will simply be CHEAPER to own an electric vehicle.

What is CALSTART Doing Working to create project proposal teams CARB and CEC grants DOT and DOE grants Working with US Army TARDEC Cybersecurity Truck & Bus working group, ISAO Connected and Autonomous Truck & Bus working group; technology and users Education about technologies and pathways Focusing on emissions and VMT benefits