Why there's such a buzz about Hive.

 

Meet Nazma and learn how Hive is helping to make every journey better. Discover why it takes a combination of data scientists, engineers and more to make sure bread and milk are always in store.

 

“I’d worked in transport before, but I was in publishing when Tesco offered me this job. What attracted me was the scale of the role and the size of their fleet. Also, a lot of big tech firms are US-based, but Tesco is headquartered here in the UK. It means you can get involved in decision-making, and that was a big plus for me. This felt like a very good place to put my energy. Tesco has a big focus on environmental impact, a subject close to my heart, so I knew I would enjoy making a difference and helping to reduce carbon emissions.

Then there’s the reach. So many people rely on Tesco stores and online for their essentials. If you're selling bread and milk but you run out, it has a big impact on people. So there’s a bit of urgency around what you’re doing. We're trying to save mileage, save trips, and still get people’s shopping to them when and where they need it. And because we have one of the largest fleets in the country – about 11,000 vehicles across trucks and vans – you have every chance of making a difference, even with the smallest percentages of optimisation.

There’s just so much opportunity. Transport is an integral part of the supply chain, so it will never stand still. One of the reasons I’m interested in transport is because vehicles are becoming more and more intelligent every day. We’re trying out different types, seeing what they can do, and what that means from a scheduling and colleague perspective. Such as how best to highlight charge data of electric vehicles to colleagues managing the operation.

Since joining five years ago, I’ve worked on two major projects in Tesco’s Product team: Hive for our truck fleet (HGVs) and Bumblebee for our van fleet. They’re both transport scheduling systems based on the same classic computer science problem: a traveling salesperson needs to deliver a series of items to a series of locations – so what’s the most efficient journey?

That's the premise of Hive and Bumblebee. Finding the most efficient and optimal route for all journeys in milliseconds. You’ve got all these orders and goods that you need to deliver to different sites. And you need to take into consideration the constraints such as the travel distance, the travel time, the vehicle type and the load sequence on the vehicle. We’re now able to cluster orders, deliver them together and do fewer miles. So, we're creating fewer emissions, which is better for the planet.

 Bumblebee kicks in when you go on our website to choose your slot for home delivery. It dynamically calculates how much capacity is in your area, based on the number of vans and other orders. Whereas Hive helps plan truck journeys the night before – pulling data from forecasting orders to forecasting sales – so the depot can start allocating journeys to drivers when they arrive the next day. This helps us plan for every scenario, like: is it going to be a hot day tomorrow and do we need to send more frozen goods into stores? So, we can better advise our transport colleagues, and they can make sure they have the capacity to deliver. And our people don't need to spend time figuring out whether or not something's feasible, because Hive is doing it.

Bumblebee began its roll out when I first joined. As we saw the numbers coming back, that’s when we came up with the hypothesis for Hive. We thought: if we can do it for vans, we can do it for trucks. We got the funding to run a 12-week discovery session where, as part of that, one of our product managers worked the night shift for a week to see how the planners were working. That’s when we realised Hive needed to be a completely separate system to Bumblebee.

I remember the moment. We were together in a room, and it took us a while to process. Then, we decided to get a data scientist to look over everything. And we knew we were going to need four scrum teams: one to tackle the routing challenge; one to tackle resource management; one to tackle the fulfilment network; and one to be the optimiser – the engine of it – which would take all that data and create a schedule. Our next step was to find the right people to help us build it.

We needed data scientists to create the algorithms. Engineers to make sure the algorithm can be coded and run – that it’s all viable and doesn’t cost the earth in hosting costs. Product managers with a maths or analytical background. Geospatial specialists to manage the road network, as well as how we source and manipulate that data. Plus, a UX team to help us understand how to get the data from end-users. Together, they’ve created UI (user interfaces) where you can configure the number of vans or trucks, or people driving. You can see council restrictions and view different routes. We’re working with real geniuses, at the top of their field. Being able to tap into different experts, dotted around the globe, was incredible.

My role in all this was about setting the vision and purpose. Keeping people motivated and tackling setbacks in the right way. I also came up with the name ‘Hive’. Because, while a bumblebee is one of the most efficient routing creatures on the planet, a hive is bigger. One of the reasons why I moved from publishing to transport is it has real-life impact. You go home knowing you’ve made big savings and put food on the shelves."

So why not explore our latest Tech opportunities? Click here