Will Artificial Intelligence push CN off the rails?

AI promises to revolutionize transportation; issuing a death sentence for railway companies like Canadian National.

Artificial intelligence (AI) is the next wave of revolutionary tech innovation, promising to re-invent every industry – from crime-stopping facial recognition, to telehealth and voice-activated shopping.  In the transportation industry, tech titans like Elon Musk, are betting on autonomous vehicles to disrupt the supply chain and bring in a new wave of fast, efficient, and safe cargo transport.1  While this is exciting, the AI technological trend also poses a significant competitive threat to established transportation companies like the Canadian National Railway (CN), whose future longevity and market share is now at risk.

Historically, railways have enjoyed a comfortable business position as the leader in moving large volumes of goods over long, land-locked distances2.  Now, however, autonomous trucks are encroaching on the railway’s positioning, promising increases in safety, efficiency, and navigation, while tempting customers to minimize or eliminate their use of rail altogether.  Autonomous trucks threaten railways in three main areas:

  • Distance: Advances in self-driving technology are enabling trucks to drive longer distances than ever before, at a faster rate, without the energy and resource constraints of human-driven vehicles.2  Given the current rates of advancement, self-driving trucks may soon outpace the time savings/distance advantages that railways previously enjoyed.

  • Safety: Self-driving trucks claim to be safer than human-operated vehicles, promising to eliminate 90% of all road traffic accidents caused by human drivers.3  Self-driving trucks rely on complex algorithms that make faster reaction-time decisions, while incorporating advanced monitoring technology to adapt to adverse weather and traffic conditions with unparalleled diligence and speed.3
  • Capital & Innovation: Railroad assets are both expensive and long-lasting, with an average fleet age of 18.4 years.2 Unfortunately, long purchasing lead times hinder the railway’s ability to respond quickly to market demands and adopt new technologies.  In contrast, commercial trucks turn over every 3-4 years, allowing for faster adoption of technological advancements.  Unless rail cars are renewed more frequently, railways will be left operating ancient machinery, thus motivating consumers to choose the faster, AI-enabled trucks of the future.

CN Rail’s Response:

As AI-powered transportation has garnered more traction, Canadian National Railway has not been idle.  CN rail recently announced that it will invest over $500M in the next five years to “improve its safety, efficiency, and competitive edge, partly in response to the potential threat posed by driverless trucks.”4  In the short term, CN will “invest in automation and strive to improve operations” to stay ahead of changes to the trucking sector.  For example, in 2017, the company plans to spend over $28M in train defect detection equipment, that will enable earlier discovery of mechanical malfunctions, thus improving safety performance.  In addition, CN has piloted the use of wearable technology such as “Readiband” to monitor and record employee fatigue and alertness to reduce the impact of human-related incidences.4


CN rail is also adjusting its competitive strategy through innovations in product shipment technology.  CN is currently testing a process called “CanaPux”, a revolutionary method to ship heavy crude from Alberta’s oilsands offshore to Asia, in efforts to establish a competitive stronghold in the transportation of crude.4  The CanaPux process turns crude oil’s raw material, bitumen, into semi-solid pucks that exhibit superior safety advantages when loaded and transported.  In the event of a spill, the pucks float to the surface and do not dissolve in water or burn, minimizing environmental impact.

In the medium to long-term, CN is betting that these significant investments in technology and infrastructure will be enough to ward off competitive threats from driverless trucks.


While CN’s investments in technology and innovative products are exciting, I believe that there is more that CN can do to protect itself against AI’s competitive threat.  If the extreme scenario occurs, where driverless trucks begin to overtake railway market share, CN is ill-equipped to defend its competitive position, particularly if the railway only makes marginal improvements in their technology.  To improve their defensive position, I propose that CN focus on two main initiatives:

  1. Tie CN’s future to the future success of AI: CN should invest in AI-based logistics companies like Otto5,6, or forge business partnerships with driverless trucking firms like Tesla,1 while establishing the CN brand as a “one-stop” shop for all ground transportation. With this strategy, CN is poised to benefit financially and strategically when the inevitable AI-based logistics revolution arrives.
  2. Continue to develop new technologies like the CanaPux that differentiate CN in the transportation market, to encourage customers to choose CN rather than the autonomous truck option.

For discussion:

  1. Advancements in AI will inevitably displace human workers. How should companies with unionized workforces (like CN) manage this trend?  What can companies do to support these workers while continuing to innovate?
  2. What should government’s role be protecting strategic assets such as logistics and transportation systems? Does this help or hurt innovation?

Word Count 791


  1. Goodkind, N. (2017, November 14). Elon Musk reveals Tesla Semi truck that will “blow your mind out of your skull and into an alternate dimension”. Retrieved November 15, 2017, from http://www.newsweek.com/tesla-semi-truck-elon-musk-710689
  2. Oliver Wyman. (n.d.). Disruption: The Future Of Rail Freight. Retrieved November 15, 2017, from http://www.oliverwyman.com/our-expertise/insights/2017/sep/oliver-wyman-transport-and-logistics-2017/operations/disruption-the-future-of-rail-freight.html
  3. DHL Trend Research. Self Driving Vehicles in Logistics. 15 Dec. 2014. Retrieved Nov 15 2017, from, https://delivering-tomorrow.com/wp content/uploads/2015/08/dhl_self_driving_vehicles.pdf
  4. “CN to spend $500-Million in tech with eye on driverless truck threat.” The Globe and Mail, 14 June 2017, http://beta.theglobeandmail.com/report-on-business/cn-to-spend-500-million-in-tech-with-eye-on-driverless-truck-threat/article35309980/?ref=http%3A%2F%2Fwww.theglobeandmail.com&
  5. “AGV VS. SDV; A comparison of automated material transport.” OTTO Motors, ottomotors.com/resources/info/agv-vs-sdv.
  6. “In 10 Years, Artificial Intelligence Will Transform Trucking, Says Otto Exec.” Fortune, https://fortune.com/2017/03/27/uber-otto-artificial-intelligence-truck-driving/.


The Race to ROI: Can Adidas’s Speedfactory Make Hyper-Customized Sneakers and a Profit?


Will Johnson & Johnson continue to be a climate change champion as it becomes more global?

Student comments on Will Artificial Intelligence push CN off the rails?

  1. I found this article really interesting – especially the discussion questions as I’ve come from a unionized, highly manual organization before. The pessimistic side of me believe that at a certain point, AI will essentially be able to perform every job, which is why many are talking about a future universal wage if enough workers get displaced. In the short term, companies will need to start making the AI decision on a case-by-case basis (as their entire workforce won’t necessarily be replaced at once). In that case, I think the best defense will be skills-retraining for jobs that have not yet been replaced. It’s unlikely that unions will like any of the options that mean losing jobs, I believe that’s the best solution (helping them find new ones).

  2. I enjoyed reading your essay, Olivia, and found the topic interesting. One question that stands out in my mind is: why can’t Canadian National also develop AI to produce “driverless trains,” just like driverless trucks. For the agriculture industry, innovative companies are developing driverless tractors that can plant and harvest fields based on set parameters. (This is an interesting video on driverless tractors: https://www.digitaltrends.com/cool-tech/self-driving-tractors/). It would seem that trains would be able to be programmed through parameter inputs similar to driverless tractors and trucks. By developing driverless trains, CN could be able to realize some of the same improvements in safety and efficiency. I did not see any information mentioned in your essay regarding why CN has not pursued this option. Are there barriers that would prevent driverless trains from being developed?

    CN has been very innovative in their short-term strategic response with the develop of train defect detection equipment and “Readibands.” It would seem that these innovations could be utilized in many industries. If CN was able to patent the technology, they might also be able to monetize the innovation through sales of the technology for many different applications.

  3. Great idea looking at AI on a lower-profile transport method! Government shouldn’t step in to protect this strategic asset, because I don’t believe autonomous trucks can replace trains. I agree that AI will potentially eliminate labor costs for the trucking industry, making it more competitive. But I see the crucial issue not around labor, but rather fuel. Trains are perfect for low-value, high-weight, and less-time-sensitive items (often raw materials). Shipping sand or iron ore, for example, is too costly using vehicles. Government should allow the space for the rail industry to respond to autonomous trucks. They may in fact be highly complementary: autonomous vehicles are perfect for “last mile” delivery that trains can’t handle. Governments can justify the need to protect strategic systems, but I don’t think they should protect the rail industry just yet.

  4. Hey Olivia, really cool topic and great write-up! It will be really interesting to see how it all plays out, but one question I have regards Tesla’s development efforts: are they proposing driveless electric trucks? If so, my impression has always been that the battery costs associated with electric trucks would make it difficult for them to compete on cost with rail. That being said, I think there’s significant opportunity for innovation here, since I would suspect rail to be a much easier transportation platform to automate than vehicles. CN could definitely aim to lead the way on this front.

  5. Fascinating read, Olivia! AI will definitely, displace workers; however, given cargo loads can be valued in the hundreds of millions of dollars, CN and other rail companies may still require a crew on their locomotives even once they adapt more AI to “drive” (to manage the train in the event of emergency stops, etc.). Would be really interesting to see if there is an AI push in terms of safety measures given what happened at Lac Magantic in 2013. The CanaPux is a fascinating idea considering the challenges with building pipelines in BC to access asian markets for oil sands products. Lastly, perhaps CN would partner with certain AI logistics companies for “last mile” delivery of goods.

Leave a comment