Research Engineer

About us

Hypermile is on a mission to improve trucking efficiency, sustainability and safety by developing an AI driver assistance system 🚛 🌎🌱.

Today, fuel is the main cost component for logistics operators, accounting for >1/3 of all operating costs. Driving behavior is a key factor affecting overall fuel consumption and can increase fuel use by up to 35%. However, current solutions are not cost effective, so logistics operators are constantly looking for new solutions.

In addition, logistics operators are under huge sustainability pressure, as Heavy Goods Vehicles (HGVs) are responsible for 16% of the UK's transport emissions but only 5% of the miles travelled. While electric vehicle sales for passenger vehicles are on the rise, the same can't be said for long-haul HGVs due to high initial cost and limited range, making it impractical for logistic purposes.

With this problem in mind, we aim to leverage the latest AI techniques to improve the fuel efficiency of diesel trucks today, and increase the range of zero-emission trucks of the future.

Our product Hypermile Co-Pilot is a retrofittable AI cruise control for commercial vehicles focused on controlling the speed of the truck efficiently. We have trained the AI algorithms to learn the best techniques for saving fuel: anticipating how traffic flow will change, optimising speed based on the road gradient and maximizing vehicle coasting. As a result, we can reduce diesel consumption by 11% and extend the range of battery-electric vehicles by 15%. The long term vision is to incrementally build higher levels of autonomy with the next milestone being a Level-3 autonomous trucking solution.

The company was founded in May 2020 and raised a $1.5M pre-seed round in May 2021 from Y Combinator, Greg Brockman (CTO of OpenAI), Luc Vincent (former Executive VP of Autonomous Driving at Lyft Level 5) and others. We’ve received a number of R&D grants from Innovate UK, Department for Transport and Horizons 2020. We’re currently a team of 10 across Commercial, Product and Engineering (including 4. sub-teams cloud software, machine learning, automotive software and electronics).

What you’ll be doing:

You will be part of the machine learning team who are responsible for training, deploying and monitoring the performance of machine learning models at the company. This primarily involves computer vision models running on our edge-computing device. Tasks will include designing environments for training and evaluating robust and efficient computer vision models, integrating and building upon recent research advances to produce new IP, and optimising inference models for specific chipsets. The ideal candidate will have experience in delivering production computer vision/machine learning systems and a solid background in software engineering and computer vision/machine learning.

This role is a UK-based hybrid-working role where we co-locate in London and Millbrook Proving Ground. Candidates must have the right to work in the UK.

What you must have:

• BS or MS in Computer Science or related field.
• Experience in designing and deploying machine learning and/or computer vision models from start to finish.
• Experience working with modern open-source machine learning frameworks e.g. PyTorch, TensorFlow.
• Experience working in production software engineering teams so are familiar with at least some of GIT, AWS/GCP/Azure, Terraform.
• Experience working on at least one of:
  • End-to-end machine learning or reinforcement learning projects.
  • 3D computer graphics engines or simulation tools e.g. Unreal Engine.
  • Computer vision models for ADAS/autonomous vehicles.
  • Optimising deep learning models for specific chipsets (e.g. Nvidia, Qualcomm).
• Knowledge of 3D computer vision and computer graphics.

Bonus points!

• MSc or a PhD in Machine Learning, Computer Vision or related subject.
• Track record of publications and/or patents in machine learning or computer vision.
• Experience in developing and maintaining large-scale dataset workflows.

Don’t meet all of the role requirements? Don’t worry! We care about outcomes rather than tickbox criteria, and it’s often the case that the perfect person for the role has a CV that looks a little different. Research (by Harvard University) shows that women are particularly likely to second-guess themselves and not apply — so if you’re worried you don’t meet all the criteria, get in touch anyhow and let us do the worrying.

Why Hypermile?

• Shape the company. We are very inclusive in our decision-making and very committed to having the best idea win.
• Grow with us. We may be a young company, but we already have big plans and are aggressively executing with us. We are very keen to enable everyone’s growth—whether you want to be an engineering manager or a world-leading engineer.
• US-style stock options. When we succeed, we want to take you along for the ride!
• 28 days paid holidays. We believe having time to recharge is important.
• Flexible Working. We offer flexible hybrid working allowing people to work when and where works for them.