Request the latest Journey Intelligence COVID-19 insights

By clicking submit you consent to sharing your data with wejo for the purposes of contacting you regarding wejo’s products and services

Your request for Journey Intelligence COVID-19 insights has been submitted, we will be in touch shortly.
Something went wrong while submitting the form.
wejo

This content is best viewed on desktop

US-wide traffic is starting to decrease

States are reacting with increased restrictions on businesses as the coronavirus pandemic continues to impact Americans, resulting in a drop in traffic across the US

Data science overview

The rate of growth in miles travelled across the US had been increasing week on week since mid-April. The trend has now started to reverse as cases of COVID-19 infection start to rise, we see this reflected as a 2% drop in traffic compared with the previous week.

Changes in traffic volumes for individual states follow broadly the same pattern, with some notable exceptions.

Texas, one of the worst hit states for the coronavirus pandemic in the US, saw travel decrease by 8% compared to the previous week as the state tightens restrictions on businesses again. Florida and Arizona also saw a decrease of 2.9% and 4.5% respectively as the pandemic continues to hold a grip over their communities.

There are states where traffic has continued to increase, these include New Jersey (3.6%) and Virginia (4.1%). This is widening the gap between states where traffic volumes are continuing to increase compared to those where traffic volumes are dropping.

Traffic volume forecast

Using actual volumes of data for the first 5 months of 2020 to estimate post-lockdown traffic volumes

More than two weeks into our June traffic predictions and we can validate our results, as well as update our predictions for the next 30 days as we continue to see travel patterns move closer to a pre-coronavirus state.Our traffic volume predictions during the first half of June were accurate to within 6% of actual traffic volumes. Based on these results, we updated our traffic prediction algorithm and re-calculated our forecast for the remainder of June. We predict traffic will increase by 0.9%, down 5% compared to pre-COVID volumes.

This is predicted with an 80% confidence level. Our new 30-day forecast shows that as we move into July, traffic will follow a similar upward trend which will remain steady with no signs of tailing off. However as we come closer to pre-coronavirus levels, will this rate of growth slow?

Traffic forecast: background and original predictions

Data science overview

Throughout the course of the COVID-19 pandemic we analysed the impact it has had on driving and journey trends. With the US now relaxing stay-at-home orders and our data showing an increase in car journeys, we wanted to explore whether it was possible to predict traffic trends. Using traffic volume data from January to May 2020, we created a 30-day traffic forecast that estimates that during the last week of June, traffic will be around 4-8% lower across the US than in the first week of March (pre COVID-19). Even as the US moves towards a complete lift on lockdown restrictions, car usage might not return to normal while there is still an outbreak.

Comparing journey behaviour to current day US

March - May 2020

While our forecast was calculated based on the current trend of steady increases in traffic, lockdown restrictions aren’t the only factor impacting traffic levels. There are a number of external factors – from changes in working habits to shopping trends - that make it challenging to predict what will actually happen and create a lot of uncertainty. We will track our forecast against actual traffic volumes to see how much it deviates and why.

Volume of journeys to Grand Canyon National Park

March - May 2020

Data science overview

Like many parks across America, the Grand Canyon National Park closed its gates to the public on 1 April following stay-at-home orders to reduce the spread of coronavirus. As the US - along with the rest of the world - starts to re-open businesses and resume some sense of normality, the Grand Canyon opened up access to its South Rim from 15 May.

Our data reveals that even before the restrictions came in on 1 April, visitors to the Grand Canyon started to decline significantly. We then see a plateau of a limited number of visitors to the park throughout April. The Park announced on 13 May that it would reopen two days later, and we see in our data that visits start to climb in the days leading up to this.

Volume of journeys to Bryce Canyon National Park

February - May 2020

Data science overview

The United States is continuing to take measures to ease lockdown restrictions to allow certain businesses and organisations to re-open. Amongst those organisations are parks and recreation grounds and centres. Bryce Canyon National Park, in south eastern Utah, draws more than 1.5 million visitors every year, but has been closed since 7 April to reduce the spread of coronavirus. It was widely reported that it was reopening from 6 May to enable visitors to return to the park to hike and enjoy the scenery.

To provide an insight into the impact of stay-at-home orders and easing of restrictions on visitor behaviour, we analysed visitor volumes and aggregated and anonymised journey start locations of those travelling to Bryce Canyon from the end of February - to benchmark behaviour before COVID-19 - right up until the beginning of May.

Our connected car data reveals that pre-COVID 19, visitors came from a broad catchment area, from Las Vegas right up to Salt Lake City. Throughout March, as coronavirus continues to spread across the US and in the lead up to the park’s closure, we see visitor numbers start to decline, with visitors typically coming from local areas only.

From 6 May, our data shows a steady increase in the volume of visitors and we start to see to see visitors come from further afield, in a similar pattern to that seen pre-COVID 19.

Where did people in Georgia drive to for their first journey in a two-week period?

1st - 2nd May 2020

Data science overview

Georgia was the first US state to ease lockdown, with businesses including gyms, hairdressers, tattoo studios and restaurants all now able to reopen their doors to some degree. To provide context to our analysis on travel patterns following the easing of restrictions, we looked back to pre-coronavirus trends.

Our data science team analysed vehicle usage on Fridays for February, March and April. In Georgia in February, vehicles were stationary for around 15.5 hours. Vehicle usage decreased in parallel with the coronavirus pandemic and shelter-in-place orders. On Good Friday (10 April), vehicles were stationary for 20% longer than in February, marking the peak of the downward trend in vehicle usage.

Broadening out the analysis to surrounding southern states – including South Carolina, Mississippi and Tennessee – we see a similar trend emerge of vehicle usage gradually declining up until Good Friday. Following the Easter weekend, vehicle usage starts to increase per 24-hour periods in Georgia and its surrounding states - despite stay-at-home orders being in place at this time.

Georgia provides a relevant case study as to how the easing of lockdown could impact the surrounding southern states, and beyond. We analysed the first journeys made where vehicles hadn’t been active for at least a two-week period.

Despite the opening of beauty salons and restaurants, etc... we can see that the majority of journeys were either a round trip – likely to a drive thru – or to visit friends of family. Only 2.8% visited hair and beauty establishments, while sports and recreational trips made up just 5.7% of journeys.

We will continue to analyse and report on trends in Georgia and the surrounding states.