Many transportation companies recognize AI as a game-changing technology, however some barriers to adoption still exist. As adoption of artificial intelligence (AI) in the transportation industry continues to slowly grow, Gartner recommends that transportation CIOs follow a four-step process to ensure success with AI projects.
"The current environment has triggered low ridership numbers and operational disruption in the transportation sector, prompting a major need for cost optimization amongst transportation companies," said Pedro Pacheco, Senior Research Director at Gartner. "Some organizations have implemented AI to help drive cost savings, however that number remains low. According to a Gartner survey, only 12% of transportation respondents have adopted AI, while 35% of respondents of other sectors implemented AI to deliver greater cost reductions or a larger revenue increase."
During Gartner IT Symposium/Xpo EMEA, Gartner analysts explained the most common pitfalls transportation companies encounter when deploying AI. Demonstrating the industry applicability and business benefits of AI and integrating it into existing infrastructure are the key obstacles to AI progress in the transportation sector.
Gartner identified four steps that can help transportation CIOs better formulate use cases and identify which business process operations can be extended to AI capabilities:
Step 1: Define AI Use Cases
Transportation CIOs should start by gaining a broad understanding of all areas and use cases where AI can be used in their organization to achieve financial benefits. Essentially, AI must be seen as a tool to solve problems in areas that are critical for transportation companies, such as the reduction of operational costs or greater customer centricity like service reliability and speed of delivery.
"Transportation CIOs should work closely with several experts and vendors who can help them identify specific AI use cases and quantify their benefits," said Pacheco.
Step 2: Create an AI Growth Plan
Transportation CIOs should develop a concrete plan that demonstrates how AI can help the company achieve growth — either by setting specific revenue targets or through cost optimization that ultimately frees capital to make other investments.
The growth plan should also focus on both short payback period actions and long-term deliverables. Some companies … created entire standalone AI teams that not only develop AI applications but educate their organizations on the benefits of AI. This approach enables a greater variety of AI applications across the entire organization.
Step 3: Present Plan to the Board of Directors
Once the AI growth plan is finalized, transportation CIOs should present it to the board of directors. In many cases, this plan will imply adapting the company's long-term strategy to account for the benefit of AI, along with the necessary investment.
As such, showing both long-term growth opportunities and short-term wins to the board of directors is important.
To address possible initial skepticism, ensure several case studies that demonstrate successful AI projects are included.
Step 4: Gather Resources
In order to progress and complete an approved AI growth plan, transportation CIOs need to garner the right resources. Gartner analysts said that if the right AI skills and capabilities do not exist internally, talent can be acquired, or external partners can be selected to help the organization acquire an AI skillset.
"Developing successful AI applications requires a multidisciplinary team that gathers both IT expertise and business process owners. This is critical to validate that business requirements are carefully set as well as to ensure the AI solution will tackle the business problem," said Pacheco. "Companies that are serious about AI need to embrace it as a company-wide priority rather than relegate it as an IT-only objective."
Methodology: Gartner survey was conducted online during November and December 2019 among 607 respondents from organizations in the U.S., Germany and U.K. to uncover the factors contributing to the success of AI implementations and the main barriers of AI operationalization.
