London-based startup has developed a cutting edge machine learning algorithm that predicts parking occupancies. The company is offering its services to parking manager or a mobile parking to analyse parking payment data and to generate predictions of future occupancies in order to implement a dynamic pricing system based on yield management. As a result, an optimal use of parking resources will be obtained, together with an increase in revenues.
The London-based startup has developed a Bayesian machine learning algorithm that predicts parking occupancies from payment data. Among the possible applications, these predictions can be used to implement a dynamic pricing system, where fees are chosen taking into account the expected demand. Taking as input as little as two months of past payment data, an accurate prediction of the parking occupancies for the coming week can be achieved. The use of payment data, instead of real time occupancy data, prevents the need for cameras and sensors, which have a high cost of both implementation and maintenance. The company is seeking a parking manager or a mobile parking application willing to share the parking payment information in their possession and implement a dynamic pricing strategy based on the company's predictive algorithm. This does not require any initial investment and the ensuing increase in revenues will cover the company's fee. Dynamic pricing is just one of the possible application of the predictive power of the company's algorithm. Others include: - The predictions themselves can be showcased to drivers to facilitate the planning of their journeys; - Knowledge of future occupancies can also be used to optimise any booking system.
Type (e.g. company, R&D institution…), field of industry and Role of Partner Sought:
The company is seeking a parking owner or a mobile parking applications to embark in a proof of concept project, with the aim of implementing the company's dynamic pricing system, which in turn is based on the company's predictive algorithm applied to parking payment data. The partner will provide the appropriate parking payment data to the company, which will then use its state of the art algorithm to analyse said data. Each week the company will generate a pricing scheme for the coming week. The partner will then have to implement the recommended pricing fees, the use of which will guarantee an optimised revenue stream. From the success of the proof of concept stage, the company expects to obtain a more long-term work relationship. In order to access the adaptive optimal pricing strategy delivered by the company, the partner will pay a monthly fee which could be either fixed or flexible and connected to the number of drivers using the parking facility. Moving forward, other applications of the algorithm can also be considered, together with other data analysis services.
Technical Specification or Expertise Sought:
The company is looking for commercial partners operating in the parking industry to implement the company’s technology to accurately predict parking demand. In order to implement the services offered by the company, the partner will have to provide parking payment data. Specifically, the parking manager or mobile parking application will provide time of arrival and amount paid for parking.