<p>An Intelligent Transportation System (ITS) requires scalable and fault tolerant computation models to process the large scale data of traffic sensors. The design of a more generic multi-modal ITS is relatively challenging because the traffic information is computed using multiple data sources, many of which are not primarily deployed for traffic monitoring. The ITS should compute and provide traffic information to the traffic applications as per their granularity and timeliness requirements. Further, the storage for traffic data should be secure and immutable. To address these requirements, this paper proposes a distributed computing and secure storage framework for multi-modal ITS. The framework anonymizes user specific data, represents ITS data processing tasks as the distributed processing and aggregation operations, and utilizes a blockchain based distributed ledger for storing the traffic information. The framework provides scalable, fault tolerant, and privacy preserving data processing, and the fault tolerant and immutable storage for traffic information. We present a case study of designing a multi-modal ITS using the proposed framework. The multi-modal ITS processes the raw data of mobile phone users, Global Positioning System (GPS) enabled probe vehicles and the selectively deployed traffic sensors for generating traffic information. The MapReduce framework is employed for the distributed processing of traffic data, and the Ethereum blockchain is used for storing the traffic information. The communication and storage overhead of distributed processing, the computational speed-up gained through the framework, and the resource requirements to serve the traffic applications are analyzed for the road networks of different sizes. The blockchain access latency is evaluated using the Rinkeby Ethereum testnet and local blockchain. The results are encouraging and exhibit the feasibility of using the proposed framework for a large scale ITS deployment.</p>