[1]<\/sup><\/a>.<\/p>\n\n\n\nThis shortage created a good opportunity for self-driving car\nmakers, to navigate the market and seek the opportunity to manufacture\nself-driving trucks, that can fill the gap and introduce new products to the\nmarket, that will allow moving trucks on the highways with limited interactions\nfrom drivers or without a driver, by building full automated self-driving\ntrucks (Halsey 2017). Self-driving trucks are already being tested in\nthe U.S., in 2016, Otto\u2019s first self-driving truck successfully delivered a\nshipment of beer on a 120-mile trip, while the driver is setting in the\nbackseat and observing the behaviour of the automated truck (Davies 2016).<\/p>\n\n\n\n
The self-driving vehicles\nindustry in general is on the rise, with 44 companies working on building their\nown autonomous cars (Cbinsights 2017), meanwhile companies like Daimler, Otto,\nTesla and Volvo are working on producing new self-driving trucks with different\nlevels of automations that range between assisting the driver to increase their\nsafety, to a full autonomous truck (Muoio 2017). The American Society of Automotive Engineers identified six different\nlevels of vehicle automation, starting with zero which has no automation to\nfive with full automation (SAE International 2016). Nowadays, the market offers\ntrucks at level one with limited driver assistance technology, and many\nresearch and development efforts are happening to build and introduce new\ntechnologies, hoping that will produce higher level of automated trucks,\nintroducing new market and regulation challenges.There is no big difference between the required technology to\nbuild an autonomous car or a truck, they are both relying on advanced machine\nlearning, sensors and communication systems as the base of a complex system\nthat will make the care make decisions on the road. On the other hand,\ndifferences will start to emerge between both products when assessing the\npotential markets and trying to understand the legislative frameworks for both\nthe business and transportation laws and regulations (Halsey 2017).<\/p>\n\n\n\n
For the past seven years, 29 U.S states passed their own laws and\nregulations that govern self-driving vehicles. In 2017, the U.S Congress passed\nthe (SELF DRIVE Act<\/em>), to cover the\nself-driving cars regulations on the federal level. However, due to the\ncomplexity of the issue, and the pushback from the unions, the act didn\u2019t\nregulate the self-driving trucks, and only passed regulations related to\npersonal cars (\u201cSelf-Driving Vehicles\nEnacted Legislation\u201d 2019).<\/p>\n\n\n\nThe Truck industry in the U.S sees self-driving\ntrucks as a threat to the 3.5 million truck drivers (Hassler 2017) which\nalready mobilized drivers and unions to advocate against autonomous trucks. On\nthe other hand, a more possibly damaging cyber threat is growing, in parallel with the rise of computer\ncontrols and digital communications over cars, and several proofs of concepts took place in the\nlast couple years (Tuttle 2017). beside all of that, there will be a need to improve the infrastructure of roads and\nhighways before allowing trucks on the road. The market of self-driving trucks is promising.\nCompanies like Uber, Amazon, USPS, UPS and PepsiCo are funding R&D and\nper-ordering trucks from manufacturers that announced plans to produce\nself-driving trucks.<\/p>\n\n\n\n
Technology\nroadmap<\/strong><\/p>\n\n\n\nTo create a roadmap that captures the status of\nself-driving trucks and maps the possibilities and challenges of its future, flexibility is needed.\nHence, roadmaps are flexible when it comes to their format,\npurpose and use (Phaal et al 2004). This flexibility will allow adopting and\nsomehow customizing the roadmap to reflect the best visualization of\nself-driving trucks evaluation process.<\/p>\n\n\n\n
The\nfirst layer of the map is the Level of\nAutomation<\/em> which represent milestones and eras of self-driving trucks, in\nother words \u201clevels of vehicle automation\u201d according to SAE International. Bars\nwere used to present the level of automation, and to show the available\ntechnologies at a specific time. <\/strong>The end goal of this map is the actual product\nwhich is placed on the second layer of the roadmap, to\nunderstand the technology evaluation, and\nmarket and legislative are at the lower\nlayers. <\/p>\n\n\n\n <\/figure><\/div>\n\n\n\nFigure.\n1. Self-driving Trucks RoadMap<\/em> <\/p>\n\n\n\nThe\nmapping process for Self-driving trucks has different current and future\nproduct characteristics, ranging from partly automated to fully automated\nproducts. Capturing the evaluation of the involved technologies and the\nchallenges\nare represented in multiple layers or format (a) (Phaal et al 2004)\nthat can be expanded and has sublayers roadmap.<\/p>\n\n\n\n
Figure.\n1<\/em><\/strong>. Shows a multiple-layers roadmap for self-driving trucks that was\ndesigned after an extensive research on self-driving history, current status\nand its future. This can be seen in the \u2018Level\nof Automation<\/em>\u2019 or milestones layer that shows different eras of\nself-driving, starting with the current level that products are being\nmanufactured at \u2018Driver Assistance\u2019 which by itself represent a different level\nof truck\nautonomous that. For example, Ford F150\ntruck<\/em>, was the first truck introduced (Truck Adaptive Cruise Control System) which\nrepresent the least advanced technology in driver assistance level compared to Volvo Refuse Truck<\/em> which can move the\nsteering wheel and make some decisions, as the most advanced in the driver assistance\nlevel.<\/p>\n\n\n\nAlthough\nthe technology is advanced today, and several tests took place on the streets for\nauto-driving trucks – which can be seen on the roadmap – other challenges are\nslowing the technology readiness and prevent the process from moving forward to\nan advanced era of automation. These challenges can be found in at least one out of\nfour domains as Heslop et al. (2001) suggested, which can shape the strength of the technology<\/em>.<\/em> For example, the concept of self-driving\nin general is a new one, and it\u2019s mostly relying on machine cumulative learning,<\/em> which is still elementary and limited\nto only handle easy driving tasks, which puts technology builders in front of the learning\nchallenge, which takes time and effort. On the other hand, sensors\nare still not advanced enough to function during bad weather. <\/p>\n\n\n\nThe\nmarket of self-driving trucks is the most important element in the process,\nwhich is not the case in self-driving cars as the decision at the end will be\npersonal compared to\nself-driving trucks that needs a market approval, adoption and flexibility from\ntrucks manufacturers and future clients. This can be done by capturing how the market\nis engaging with the topic to comprehend the size of investment that will be\nput in R&D at this stage and the production in the future. The map captures\nthe \u2018The Market Attractiveness\u2019<\/em>\n(Heslop et al. 2001) including current and future deals and how such deals and\ndemand are pushing the development of self-driving trucks.<\/p>\n\n\n\nGovernments\nhave power over businesses and manufacturers and they regulate transportation\nwhich puts the legal\/legislative aspect of self-driving trucks in a very\nessential position. Especially, that there are already many loud voices against\nregulating self-driving trucks, mainly fuelled by fear of drivers losing their jobs and cyber\nsecurity threats.<\/p>\n\n\n\n
How the\ngovernment is dealing with the issue and how new government bodies are being\nformed for this purpose, can impact the technology and future products. Therefore,\nplacing them on the roadmap can show the direct impact on the final product and\nits future development.<\/p>\n\n\n\n
Key challenges and recommendation for the\ntechnology <\/strong><\/p>\n\n\n\nAdvocacy,\n<\/strong>Public acceptance for this technology is needed, and this can be\nachieved by answering people\u2019s related fears to self-driving trucks. This can\nbe done by protecting drivers\u2019 jobs at least for the next 10 years by adopting\nregulations that prevents any truck from moving without a driver, even if the driver is not\nactually driving the car. Such action can ease the pressure on the technology\nmakers, so they can focus on achieving their goals. <\/li><\/ul>\n\n\n\nAlso,\nincrease the amount of safety research that can clear the fear of auto-driving\nmachines in\nthe public eye, by proving its capabilities of being safer than human drivers. These\nissues can be found on the map, ongoing safety research and no trucks without\ndriver on the street regulation.<\/p>\n\n\n\n
Legal\nProtection, <\/strong>Currently, the U.S government is avoiding regulating Self-Driving trucks\nunder the pressure of Truck Drivers Unions and other lobbying groups opposing\nthe technology, this was obvious in the SELF DRIVE Act 2017, which didn\u2019t\ninclude any self-driving trucks regulations. But at the same time, the\ngovernment is not preventing developing such technology or even testing it on\nthe roads in most states.<\/li><\/ul>\n\n\n\nFor\nSelf-Driving Trucks industry to move forward, a \u2018Legal protection for\nautonomous trucking\u2019 is needed to allow the market to move forward. At the same time,\nmore research, tests and transparency are needed from the manufacturers to allow making better\ndecisions on both the government and public level.<\/p>\n\n\n\n
More\ndata is needed, <\/strong>Machine learning systems need more data to offer quality results.\nCurrently, labs rely mostly on simulations to generate data that can improve\nthe AI part of the self-driving truck (Stewart 2017). However, to collect an\nout-of-lab data, an addons technology is needed to be installed on regular\ntrucks to watch drivers\u2019 behaviour and collect information. At the same time,\nthis will give drivers the opportunity to be part of the evaluation process, instead\nof being excluded from process which they fear the most. This technology is\nreflected on the roadmap at the technology level in both knowledge information\nsystem and drivers improving machine learning.<\/li>Cyber\nSecurity<\/strong> is one of the mandates of the new U.S auto-driving legislation, under \u201cCybersecurity\nof Automated Driving Systems\u201d which prevents manufactures of selling any\nauto-driving car without Cyber Security policy and a system that is capable of\nidentifying, assessing, and mitigating any cyber-attack and prevent or correct\nin case an attempt took place (SELF DRIVE Act 2017). Same rules are going to be applied on\ntrucks in case trucks were added to the Self-Drive act in the future. It\u2019s important for the\nfuture of the technology to be ahead of the game by establishing new business model\nof threat intelligence that focus on self-driving cars, and studies\nattack techniques, malicious codes, vulnerabilities and auditing processes. The\nroadmap represents this issue in two different points, Threat Intel Cybersecurity Centre<\/em> and The Era of Cybersecurity Technology<\/em> in Design<\/em> which is getting cybersecurity experts to be more\ninvolved in the designing stage, to build secure technology rather than investing on securing it\nafterwards. <\/li>Improved\nVehicle-to-vehicle communication system, <\/strong>In December 2016, the\nObama administration proposed a mandate to add Vehicle to Vehicle Technology\n(V2V) to new cars which will allow cars to talk to each other wirelessly and share\ninstant data aiming to prevent traffic accidents (NHTSA 2016). Such technology can be essential\nto self-driving cars\/trucks. However, the proposed mandate was quietly put\naside by Trump administration recently (Lowy 2017). V2V is an important part\nof trucks automation process, and by having government making it a mandate is a good opportunity\nto increase V2V within larger number of cars, and this will make roads\nsafer and more information will be flying over the highways. The issue can be\nfound on the technology level of the roadmap, Improved V2V system<\/em>, also by mandating\nV2V system by the government<\/em>.<\/li>Autopilot<\/strong> is\nwhen a truck is fully capable of driving under a fully autonomous mode, with or without\nsupervision from\na driver.\nIn this stage, technology will be very advanced to the point that minimum\ninteraction from humans is needed. At the same time, humans can start building trust in the\nmachine by watching and using it. The roadmap represents this issue thru the following points, capturing more Data,<\/em> that will improve\nthe AI level and increase the safety and reliability of the trucks. New insurance Market<\/em> will be established\nto match expectations of future clients and the market changes. Also, the\ngovernment will be Introducing Autonomous\nVehicles Lanes<\/em> that will restricted self-driving vehicles to specific lanes, to assure\nregular drivers their safety and give them at the same time the opportunity to witness how\nautomated cars are\ndriving on the roads. This will also be beneficial to stakeholders by giving\nthem more accurate and tested data on automated cars\u2019 behaviour.<\/li>Self-driving\ntrucks are superior to average human drivers<\/strong>, this can be the goal of\nthis technology, and if self-driving industry was able to achieve it then\nself-driving trucks will become fully autonomous. In their \u2018Autonomous Vehicle Technology Guide for\nPolicymakers\u2019<\/em>, Rand Corporation, recommended that full autonomous cars can\nbe permitted only if they superior to average human drivers (Anderson et al\n2016). This was presented on the map as a level of automation or milestone that\ncan announce an era of fully autonomous trucks. \n<\/li> <\/strong><\/li><\/ul>\n\n\n\nConclusion <\/strong><\/p>\n\n\n\nSolving\nthe technology challenges of Self-Driving trucks might not be a hard task\ncompared to the other legal and advocacy challenges. Governments (Federal and\nstate) need to understand the technology and its regulatory aspects to produce\nthe needed legal protection. On the other hand, a deal is needed with Truck\ndrivers\u2019 unions and organizations, in order to include them in the process of\nbuilding this technology, as there are tools that can guarantee their jobs for\nthe near future, and that the main goal of this technology is actually to close\nthe gap between the market and the demand, which is not going away anytime soon<\/p>\n\n\n\n
References<\/strong><\/p>\n\n\n\nHalsey, A.\n(2017) As the era of driverless cars looms, can self-driving trucks be far\nbehind? – The Washington Post [Online]. Available at\nhttp:\/\/wapo.st\/2eWtOLN?tid=ss_tw-bottom&utm_term=.0701ca21face (Accessed 3 January 2018).<\/em><\/p>\n\n\n\nDavies , Alex\n(2016) Uber\u2019s Self-Driving Truck Startup Otto Makes Its First Delivery | WIRED\n[Online]. Available at\nhttps:\/\/www.wired.com\/2016\/10\/ubers-self-driving-truck-makes-first-delivery-50000-beers\/\n(Accessed 8 January 2018).<\/em><\/p>\n\n\n\nCbinsights\n(2017) 44 Corporations Working on Autonomous Vehicles [Online]. Available at\nhttps:\/\/www.cbinsights.com\/research\/autonomous-driverless-vehicles-corporations-list\/\n(Accessed 8 January 2018).<\/em><\/p>\n\n\n\nMuoio, Danielle\n(2017) Autonomous trucks by Tesla, Uber, Google will change trucking industry –\nBusiness Insider [Online]. Available at\nhttp:\/\/www.businessinsider.com\/autonomous-trucks-tesla-uber-google-2017-6\/#uber-is-pursuing-self-driving-trucks-through-otto-a-startup-the-company-acquired-last-august-but-the-project-is-at-the-center-of-a-massive-lawsuit-filed-by-waymo-googles-sister-com\n(Accessed 8 January 2018).<\/em><\/p>\n\n\n\nSAE\nInternational (2016) Taxonomy and definitions for terms related to driving\nautomation systems for on-road motor vehicles J3016.\nhttp:\/\/standards.sae.org\/j3016_201609\/ (Accessed 8 January 2018).<\/em><\/p>\n\n\n\nNCSL (2018)\nAutonomous Vehicles | Self-Driving Vehicles Enacted Legislation [Online].\nAvailable at\nhttp:\/\/www.ncsl.org\/research\/transportation\/autonomous-vehicles-self-driving-vehicles-enacted-legislation.aspx\n(Accessed 8 January 2018).<\/em><\/p>\n\n\n\nHassler, S. (2017)\n\u2018Self-driving cars and trucks are on the move [Spectral Lines]\u2019, IEEE Spectrum,\nvol. 54, no. 1, pp. 6\u20136 [Online]. DOI: 10.1109\/MSPEC.2017.7802341. Available at\nhttp:\/\/ieeexplore.ieee.org\/document\/7802341\/ (Accessed 8 January 2018).<\/em><\/p>\n\n\n\nTuttle, H.\n(2017) \u2018Risk Management \u2013 Hacking Cars\u2019, Risk Management, vol. Vol.64, no. (1),\np. p.20(6) [Online]. Available at\nhttp:\/\/www.rmmagazine.com\/2017\/02\/01\/hacking-cars\/ (Accessed 8 January 2018).<\/em><\/p>\n\n\n\nPhaal, R.,\nFarrukh, C.J.P. and Probert, D.R. (2004) \u2018Technology roadmapping \u2013 a planning\nframework for evolution and revolution\u2019, Technological Forecasting and Social\nChange vol. 71, nos 1\u20132, pp. 5\u201326. View the paper at\nhttp:\/\/www.open.ac.uk\/libraryservices\/resource\/doi:10.1016\/s0040-1625(03)00072-6 (Accessed 8 January 2018).<\/em><\/p>\n\n\n\nJack STEWART\n(2017) Why Daimler Researchers Used VR to Become Self-Driving Cars | WIRED\n[Online]. Available at\nhttps:\/\/www.wired.com\/story\/moovel-self-driving-car-experiment\/ (Accessed 10\nJanuary 2018).<\/em><\/p>\n\n\n\nLOWY, J (2017)\nAPNews \u2018Break: Gov\u2019t won\u2019t pursue talking car mandate\u2019 [Online]. Available at\nhttps:\/\/apnews.com\/9a605019eeba4ad2934741091105de42 (Accessed 10 January 2018).<\/em><\/p>\n\n\n\nAnderson, J.,\nKalra, N., Stanley, K., Sorensen, P., Samaras, C. and Oluwatola, O. (2016)\nAutonomous Vehicle Technology: A Guide for Policymakers, RAND Corporation\n[Online]. DOI: 10.7249\/RR443-2 (Accessed 10 January 2018).<\/em> <\/p>\n\n\n\n \n\n\n\n[1]<\/a> Journal, C. (2019). ATA trumpets persistence of\nshortage of qualified drivers. [online] Commercial Carrier Journal. Available\nat:\nATA trumpets persistence of shortage of qualified drivers<\/a><\/blockquote>