This page is optimized for AI. For the human-readable: Unjam traffic gym

Unjam traffic gym

Project Idea Metadata

Project Idea Description

Project idea description and market status

With the Unjam project, we aim to address the critical problem of traffic congestion and inefficiency of urban infrastructures. These inefficiencies are one of the main contributors to climate change through unnecessary CO2 emissions and cause enormous economic losses due to fuel costs and lost productivity. For context, it was estimated that 1-2% of Europe's GDP is lost annually in the 2019 report “The future of road transport Implications of automated, connected, low-carbon and shared mobility" from the Publications Office of the European Union.

Our key target customers are municipalities and companies that are tasked with planning and managing urban infrastructure as well as researchers developing cutting-edge mobility solutions. Cities, being complex and safety-critical hubs, often face significant inertia and skepticism when it comes to adopting new solutions. This skepticism stems from the gap between traffic research and ita real-world implementation, which remains a significant challenge worldwide.

To bridge this gap, the Unjam team has developed the Traffic Gym platform, which provides an intuitive yet powerful interface for interacting with high-fidelity digital twins of urban traffic networks, such as that of Zürich. The digital twins made available are calibrated with real traffic data and allow users to modify virtually every aspect of a city’s urban traffic dynamics, from the movement patterns of individual commuters to the synchronization of traffic light timings. By providing a realistic testbed, the Traffic Gym empowers researchers, transportation engineers, and urban planners to validate potential changes before implementation of novel solutions and helps in the development phase fostering exploration of new ideas. Empirical evidence from these simulations demonstrates the viability of proposed strategies, paving the way for pilot programs and eventual full-scale implementation. We believe this tool is poised to significantly aid in urban infrastructure planning and logistics, thereby alleviating traffic congestion and enhancing the efficiency of both existing and future infrastructure systems.

Currently, there is no product on the market that has the same features as Unjam’s Traffic Gym. Companies developing urban traffic interventions often propose reliable, yet standardized solutions that that do not fully exploit the specific nuances of individual transportation infrastructure systems. Moreover, only a few research groups and professionals possess advanced skills in large-scale urban mobility simulations, and those that do often lack expertise in cutting-edge areas such as machine learning, artificial intelligence, and control theory, which are the backbone of the most innovative solutions to smart mobility. Traffic Gym was developed in response to this market gap, with the aim of making sophisticated digital twins of cities accessible to all researchers and practitioners, fostering the creation of innovative solutions that can be readily implemented in the real world, thanks to our comprehensive benchmark collection developed across a diverse range of traffic scenarios. We also envision the possibility to use this tool to offer specialized consultancy services to practitioners and municipalities that require customized solutions.

Resources needed

Unjam was developed by a research team at ETH Zürich and is currently in its pre-release phase. The Traffic Gym platform has been utilized by several master’s and bachelor’s students for their theses, validating our value proposition of streamlining the process for practitioners who are new to digital twins.

We plan to use the Innovation Booster grant to make Unjam ready for market deployment, focusing on the Traffic Gym product and gather important know how to make our future start up succeed.

The project work packages are divided as follows:

The resources necessary for the above packages are listed below.