We explain how we optimize the positioning of free-floating mobility fleets
Managing a shared vehicle fleet, be it car or micromobility, is no simple task. And making sure that every vehicle is positioned in the right place at the right time, is somewhat of an art. Even Mr Da Vinci himself would marvel at the complexity and accuracy of the software being deployed in smart mobility today.
With StreetCrowd, we have refined the art of demand prediction and vehicle micro-rebalancing, providing the tools that enable operators to optimize the position of every single one of their vehicles.
Our data driven guidance is geared towards supporting operators in increasing the availability of their vehicles in the places where there is a demand for them, thus increasing the opportunity for them to be utilized. In industry jargon — maximizing utility. This enables them to better supply the market and to get the most out of their existing fleets. To achieve this a thorough understanding of demand needs to be achieved. This is not simply from observed utility of the vehicles, but based on an understanding of what users want and not how they are currently using the service.
Through working with a wide variety of vehicle sharing operators involved in car and micromobility sharing alike, we’ve found that there is no single recipe for success. Some operators require simply the intelligence on where to position their vehicles, while others prefer end-to-end operations solution. By providing all of these options, StreetCrowd contributes to the long-term success of shared mobility and, in turn, enhances the cities we live in.
Getting the basics right — intelligence on ‘hot’ and ‘cold’ zones
The foundation of the StreetCrowd service concerns getting the basics right — knowing where the demand will be. This intelligence is crucial for operators in order for them to supply users and it is based on numerous factors.
Initially, it is important to distinguish between utilization and demand. If demand is the ‘want for’ a vehicle in a certain area at a certain time, regardless of any other factors, and utilization is the actual amount of time a vehicle is in usage on a certain day, then there is some differentation here. For instance, the demand for a vehicle will not be fulfilled and thus utilized, if there are none available. Despite this, many operators base their fleet positioning based on what they believe demand will be due to their experience of the utilization of their fleets.
What this doesn’t take into account, however, is that utilization is distorted by the availability of vehicles. How would they know, for example, that an area is popular or not, if there aren’t enough vehicles close to those that may want them. Another example of this distortion can be seen with the spillover from a high demand location to a low demand one. Someone may want a vehicle in one area but cannot find one nearby. That person may walk a longer distance into a different location entirely to get a vehicle. If the operator then believes that because there has been observed utility in this specific location then this must signify demand, they have been mistaken. For this reason, demand is a far more complicated calculation than at first glance and requires multiple considerations.
When we think of demand prediction a number of obvious trends come to mind. Demand is likely to peak at rush hour during the week, while demand should sink at night when the majority of people are tucked up in bed. StreetCrowd does take this information into account, however, it is far more complex than this.
First, knowing when demand will be high does not provide any indication of where that demand will be. Having collected mobility trip data for many years, we are able to combine this with historic data from a customer’s fleet. This enables us to make accurate predictions on demand fluctuations at certain times of the day. What’s more, the accuracy of this dynamic prediction model is enhanced further using additional proxy data sources such as: Facebook events, airport schedules, weather data and national holidays. Brought together, these data sources provide a more refined indication of expected demand. Once the operator knows where the ‘hot’ (high demand) and ‘cold’ (low demand) zones are, they can get to work on micro-rebalancing their fleets by repositioning vehicles in locations where they are most likely to be used.
With dynamic demand predictions established, our AI-powered micro-balancing engine automates the identification of vehicles eligible for relocation, matching them with areas of increased booking frequency. With this, StreetCrowd enables operators to assess the potential utilization of each one of its vehicles. Imagine the following scenario: a vehicle is currently in what would be considered a “cold zone” (low demand area), however, this zone is predicted to become “hot” in a couple of hours. With this intelligence the operator can make the decision whether to relocate the vehicle, or to leave it there. The operator may want to focus on rides that will bring it the most revenue, or it may want to increase the availability of its vehicles in areas previously undersupplied.
To make this easier, we provide the tools enabling operators to simulate the utilization potential of a vehicle in a specific location at a certain time. With this, operators gain greater control over the utilization of their vehicles, allowing them to focus on targets like maximizing the number of journeys, increase availability in certain areas or to maximize revenue intake. Importantly our micro-balancing intelligence allows operators to become more agile, giving them the tools to adjust their operational strategies as they go.
AI + App for service teams
For some vehicle sharing operators, the fleet rebalancing intelligence outlined above is all they need. They access the plug-in API and use the additional insights to inform their operational and strategic decisions. Many operators, however, require a tool that can support their dedicated service teams and, in doing so, outsource a part of their operational management to us.
Equipped with our service team application, hired teams are supported in becoming more efficient and effective with their time. Instead of them repositioning vehicles that do not necessitate repositioning, or giving service teams just general predictions over demand fluctuations, StreetCrowd provides precise intelligence and the guidance needed to steer an effective service team. In this sense, StreetCrowd can reduce the involvement of the operator in ad-hoc demand fulfillment. In cases where operators are completely new to a city, having this guidance can give their operational teams a good platform to work with.
AI + App for crowd users
While some operators prefer to keep their operations in-house, with their own service teams, StreetCrowd offers a further service layer enabling crowd users to take over operational tasks themselves. In combination with existing service teams or by replacing service teams entirely, StreetCrowd’s crowdsourcing app delegates the repositioning, charging and maintenance of the vehicles to members of the crowd.
With our app, members can, for example, find nearby vehicles and move them into high demand areas. The intelligence would be the same as for service teams, but there would be the additional provision of bonifaction schemes and the handling of crowd payments. This represents a more bottom-up approach to operational management where the crowd members themselves become stakeholders in the vehicle servicing. In reality, it is more likely that operators would seek to combine the benefits of crowdsourcing tasks with a dedicated service team.
Bringing intelligence to vehicle sharing operations
How vehicle sharing operators manage their fleets goes a long way to determining the success of their service. If vehicles aren’t where users need them to be, operators will never acquire customers, nor will they keep hold of them. All the while, the revenue potential of the vehicles diminishes. Our StreetCrowd service provides the tools enabling operators to first: establish where the actual demand for their vehicles will be, and second, the guidance on where vehicles should be repositioned. This is the art of rebalancing, and at StreetCrowd we are the specialists.
Gerhard Liebmann is Head of Product at Ubiq. You can connect with him on LinkedIn.
Ubiq is shaping the future of urban mobility by enabling mobility services to become profitable. Experts in transforming raw urban data into actionable insights and valuable services, Ubiq enables better mobility decisions.