Question: what companies have you worked for

Edward Smart answered on 16 May 2025: last edited 16 May 2025 11:28

To give you a little background, I started in this role on short term contracts, so it was very much focussed on winning grants to do projects with industry. This means finding firms who want to do something to improve their business, seeing how you can help and writing a nice grant to say that if we do this, it will be really financially beneficial to the company. I.e. give us money and the return on this investment will be much larger!

So core skills I need are …
1. Searching for possible companies
2. Listening to them and seeing what their true challenges are (sometimes they focus on a symptom and not the underlying problem)
3. Checking that there is a clever bit of science, maths or AI that could address the problem
4. Writing a grant to explain to someone who isn’t in the industry that this is a really great opportunity and hopefully convincing them!

I won’t name names but here are some examples…
1. Defence company – they work on communications mainly for naval vessels but we did a project with them to use AI to monitor and predict problems on marine engines. Out at sea, if your engine fails, the ship can stop, or crash and there is no power, no light, heat, flushing toilets etc. We could predict problems with around 6 weeks lead time.
2. Dairy Filler company – they build these huge machines that take sheets of plastic and turn it into the cartons of milk you see in the supermarket. They also fill the cartons with milk, cap them and label them. We developed a novel method to detect faults. (If a fault takes place, milk is no longer sterile and must be thrown out)
3. Flight Data Company – Civil aviation (e.g. Boeing 737, Airbus 320) have hundreds of sensors on them and we can use them to check that the aircraft are being flown safely. Current approaches at the time used only a few of these sensors but I developed machine learning algorithms that could learn from many more and spot new issues. Don’t worry, aviation is VERY safe!
4. Marine Security company – this was really cool. Our coasts are rarely monitored outside of the big ports so they built this coastal sensor platform that has cameras, radar, AIS (boat transponder) and signal analysers. There are so many boats that you cannot manually track them all so we developed machine learning algorithms to flag up those vessels of interest.
5. Food Technology Company – this was a cool IT company in Scotland that was helping food manufacturers more effectively monitor their supply chains to better ensure that the ingredients they receive are trusted and accurate. We used machine learning algorithms to monitor the rate of food alerts (e.g. from the Food Standards Agency) and link them to relevant businesses so they could be alerted. It might not be a problem for them but it is something to monitor.
6. Autonomous Vessel Company – they had an idea of using machine learning to help detect a person that has gone overboard. Basically, if you go overboard, you very quickly become hard to spot in the sea, especially as night or in fog. We used a thermal camera and developed a tracking algorithm. We had a volunteer to be the ‘man overboard’ and went out off the coast to try it out. Nearly seasick, but the model worked!
7. Defence company – Submarines under the surface of the sea depend on their sensors to see where they are going and what is out there. There is no ‘car windscreen’ and even if there was, it is too dark! So they use vibration and acoustic sensors. But what happens when sensors physically fail? They may be making decisions on information that is plain wrong. So we developed machine learning models to detect when sensors have failed.

This is just a flavour. I’m very lucky to have had the opportunity to work with companies in many industries and on many problems.
The common thread is “how can we use our data to help grow our business?”
That is where I come in!