ICM Data Challenges and Needs

Common challenges were identified in the literature and in interviews with ICM system deployers. The sheer magnitude of data coming from many new and different sources presented challenges that resulted in the development of unanticipated interfaces and more customized solutions. Some of these included:

• Data architecture as an afterthought – Systems are conceptualized without consideration of data architecture and computational needs for fusion. Architectures were only vaguely described at concept and only fully articulated at the end of development. 
• Unfamiliarity with cloud systems – Agencies are concerned (and fearful) of perceived security challenges with cloud storage and use. Many agencies do not fully understand their options nor the cost implications (or possible savings) of the cloud in lieu of on-premise solutions. 
• Open source is difficult – Open source development is much harder to realize when some software developers have commercial interests to keep their solutions, data structure, or methodology proprietary. ​Occasionally, data sharing is done on the condition of omitting certain data elements. Some data providers only provide a “cleaned” version of the data rather than allowing access to the raw data. 
• The end goal for ICM is a challenge – Traveler behaviors are difficult to change dynamically.  Consequently, the jury is still out on how successful ICMs are in achieving transportation modal shift. Travel decisions are made at the beginning of commutes and once that decision is made, the travel back home typically cannot be changed (i.e., travelers that drive in the morning do not take transit home). 

Agency needs identified in the documentation and interviews conducted include the following:

• Integration of real time passenger counts – Real time transit passenger counts (e.g., from automatic passenger count systems) are missing from all ICMs. Integration of real-time passenger count data with ICM systems is essential to understanding the available capacity for transit to absorb riders shifting mode from vehicles. ​
• Data fusion guidelines – Interface standards are good, but there are few data fusion guidelines available for agencies designing and implementing ICM systems. Existing guidelines tend to be customized by corridor. Standard practices on the granularity of data and the duration of storage would be helpful.
• Decision support systems – All of the ICMs reviewed had developed decision support systems (DSS) that used algorithms to create response plans. Some ICMs had an exhaustive set of response plans to address nearly every possible condition. However, there is no guidance or standardization to date on DSS. ​
• Guidance on sharing data – Most of the ICMs reviewed indicated that agreements or MOU’s were essential to success. ​Agencies would benefit from documentation with templates and items of interest that they should consider in the development of their agreements.
• New mobility data – Most ICMs have not integrated new mobility data sources, such as transportation network companies (TNCs) or micromobility, into their systems. ​

Referencing Page: