The logistics of Coronavirus – Prof Phil Greening


Logistics has for a long time served the masters of service and cost, but now the world is changing: new mandated societal behaviours during the ongoing COVID-19 crisis suggest that a new priority of logistics should be social distancing. This is most important in the retail sector; supermarkets are now one of the biggest exposure risks for most people otherwise confined to their homes. Supermarkets are the most difficult locations for people to maintain the recommended 2 metre minimum social distance. In addition, due to recommendations encouraging shoppers to shop for what they need and not to bulk purchase, households will often visit the supermarket several times per week.

So what do we know about spreading the disease? By now it is widely understood that maintaining social distance is everything. Epidemiologists measure the infectiousness of a disease using the reproduction rate, which is a measure of how many people, on average, are subsequently infected by each infectious person. We know that the reproduction rate for Coronavirus is about 2.5 without social distancing, and estimates suggest this drops to 1.5 with social distancing. What we don’t know is exactly what the reproduction rate is if we all socially distance most of the time, but compromise when, for instance, going to the supermarket. The graph below plots the number of people infected at any one time for different reproduction rates, from a total population of 66M people.

Coronavirus Infections

Figure 1: “Flattening the curve”: the peak infection rate in a sample population of 66M becomes less pronounced with social distancing measures.

Although social distancing carries an obvious benefit of reducing the severity of the infection peak and minimising the negative impact on the health service, we can see that the flattening of the curve increases the length of time that such distancing measures may need to remain in force. This makes the case for re-thinking how we distribute food.

So how could we use logistics to flatten the curve, minimise exposure to others and therefore reduce the strain on the health service?

One solution would be to “ramp up” home delivery services by increasing the number of vans and drivers available across all companies offering home delivery services. This distribution channel currently accounts for less than 10% of the food consumed in the UK. It is not possible to acquire the number of vans required to increase this service to a level where it would have a major impact.

Another solution is to enforce social distancing in the supermarket. Regulating supermarkets in this way will have a positive impact. However, there will always be some risk associated with people travelling to and from the supermarket, passing through the pinch points, and queuing to access the building.  In a world where we are asking people to minimise their social contact should we be asking them to travel to supermarkets (possible on public transport) and queue?

However, there may be something we can learn from other forms of food distribution, particularly humanitarian logistics. In disaster situations we are often forced to replace the existing food distribution system with more effective configurations that prioritise different parameters. The main reason supermarkets exist is to provide choice and variety. These luxuries should be seen as secondary in the current crisis, the primary aim being effective social distancing to mitigate virus spread.

What if we started to make bulk deliveries to communities using articulated trucks and rely on slow, volunteer resourced final mile delivery?

We can use an example to illustrate how this might work in practice…

My village has a population of about 400 (spread over 200 households), a high proportion of which are classed as “vulnerable” by the current UK Government definition. At an estimate, the village consumes approximately one 44 tonne truck’s worth of groceries a week. The village also contains a village hall and a pub.

Under current behaviours, one person from each household in the village will visit the supermarket once a week. When there we can estimate that they will come into proximity (<2m) with, say, 6 other people. If they use public transport to get to the supermarket then that number could increase to 8. If there is a shortage of, for example, toilet rolls, it is quite likely that each household could end up making more than 1 visit per week (this may be important as the infectious period is about 5 days). This would amount to 600-800 person to person contacts a week.

If we were to collate orders in the village and place a weekly order with a supermarket (or new distribution coordinating body, such as a 3PL or volunteer group) a truck could make a single delivery to the pub or village hall. It could be unloaded by around 6 volunteers who then distribute the food to the houses over a 48 hour period; during that time they could also perform other essential duties, such as checking on the vulnerable and capturing data about who is self-isolating. Assuming that home delivery protocols (such as leaving food on the doorstep) are adhered to, this would amount to 6 person-to-person contacts a week. The potential reduction in person-person contacts is hugely significant.

Whilst the benefits may differ for other community contexts, the general principle of moving as much as possible in one movement is a fundamental rule of logistics efficiency. Furthermore, the decoupling of processes is a widely accepted approach to building resilient systems.

There may be other longer term benefits to this approach. For instance, the increased drop size gives the supermarkets an operational scale benefit, which would reduce the carbon footprint of last mile delivery: the supermarkets regain control of what gets purchased, permitting better planning, and we could maximise the benefit of testing and vaccination by prioritising the logistics resource pool.

So if we are to prioritise social distancing and self-isolation above service and costs for a substantial period of time it may be necessary to rethink logistics.