You are Home   »   News   »   View Article

3D models to improve safety

Thursday, January 17, 2013

EOS Solutions is helping oil and gas companies improve offshore safety, by modelling how people would interact with the offshore platform or equipment, like on a computer game

Most people who have worked offshore have been frustrated by badly designed systems and procedures, which made sense to someone sitting behind an office, but don't work well when put into practise.

Imagine if you could test out your procedures and equipment design before requiring staff to follow them, to see how well they would actually work?

Detroit company EOS Solutions is helping oil and gas companies to this.

It puts together computer models of equipment, and uses a computer model of people, and puts them together to see how they interact.

So you can see if people would get a repetitive strain injury doing a certain task, if they could evacuate the platform safely when they were stressed, and if they could easily find their way out of a submerged helicopter.

The computer model of the equipment can be taken either from the original engineering design models, or by doing a 3D laser scan of the actual equipment. You can do a 3D laser scan of an entire offshore oil platform.

You can do a simulation of people trying to follow the procedures you have written, to see how well they might work in real life.

The results of the simulations can generate facts and measurable data, which can be more useful in working out the right approach than opinion.

EOS does business in the oil and gas industry, aerospace and defence, and works together with Norwegian engineering company Norisol to sell its services in the North Sea oil and gas industry.

The technology was used by one oil company to see how easily crew could evacuate a platform in the event of a fire. EOS builds a model of the people, the crew, and the standard operating procedure.

'Our model took into account seasons. We can run in different weather, different sea states,' said Steve Prast, founder and managing partner of EOS Solutions, speaking at the May 22nd Finding Petroleum forum 'Beyond Macondo' in London.

Jumping out of a truck

EOS was contracted by the US Department of Defence to try to work out how to reduce soldier back injuries.

There are more injuries from soldiers in Iraq and Afghanistan from back injuries than from combat, Mr Prast said.

It took a while for the military to work out that dismounting from vehicles was a primary cause of the back injuries.

'People are transported in a relatively high truck and typically get out of it by jumping out of it. They have quite a bit of gear which adds to the weight and over time this produces repetitive trauma,' he said. 'We were contracted by the US army to look at this problem.'

The military already knew that jumping out of vehicles can hurt people, but they wanted to get a better sense of exactly what the damage was, so they could make a decision about how much money to allocate to solving the problem and other potential problems the solution might cause.

'If you talk to people in the military and tell them you're going to weld a nice stairs to the back of the truck, it's not necessarily something that they want,' he said. 'If someone's shooting at you, jumping out of the truck quickly becomes very important.'

'We found that some of the repetitive trauma injuries we see are consistent with the injuries you get jumping out of a vehicle with gear and a weapon. We could estimate the long term costs to the industry from this trauma.'

'We looked at the biometric impact, with uniforms, with weapons, exactly what does that do to a human body.'

EOS could also model how different heat conditions would change the simulation.

The company could then test different solutions and designs to fix the problem. 'We came up with a stair design that allows a quick exit but with a minimum impact,' he said.

EOS also did a training exercise looking at how easily troops could survive an Improved Explosive Device (IED).

It created a computer model of the vehicle, showing what would happen if the explosion happened in various parts of the vehicle, and whether the troops would be able to escape, and what level of physical and mental dexterity they would need.

It modelled how easily a driver could escape from a vehicle which had overturned. On the computer model, the move the driver needs to make looks like a complex gymnastic move, 'but actually he just undoes his seatbelt and slides out of there. It wasn't dramatic,' he said.

Building the simulation

You can construct a computer simulation model of any process, for example traffic flow from a city, operations inside a factory, operations on a ship or a platform, he said.

'People draw parallels between what we do and video games. It looks the same but video game people have much more money than we do,' he said.

You decompose the process into a mathematical description, and display it a screen.

'We take a system and break it down into its smallest element. Then we define what the role of that equipment is, what are its goals, what are its rules,' he said.

Then you can add a simulator of people onto a model, so the people can walk around the factory, ship or platform. 'That allows us to test plans and evaluate procedures, how in fact does that system function.'

'By defining the individual elements we build up to the system. We allow the elements to run around and interact,'

You then build a system where you can run experiments, test theories and study results.

'We modelled the operations, what does each person on the platform do, what do typical operations look like.'

Human behaviour algorithm

A key component of the simulation is the human behaviour algorithm, essentially a set of equations which show how a human being would behave in a certain situation, both mentally and physically.

The human behaviour algorithm was developed together with Trinity College Dublin (Ireland).

Physically, the person has 120 different joints which can be modelled by computer.

You can model whether someone would be able to reach something easily (for example a valve on the wall).

You can model if the person would get repetitive strain injury from doing a certain task, both over the short term and the long term (20 years).

You can model the behaviour of someone who knows what to do in disaster situations, and someone who doesn't.

You can model mental fatigue, how the person behaves differently as they get tired.

You can give people different personalities, so you can model what happens if you have 20 people together and one of them has particularly strong views about what should be done.

You can also look at what happens when people do the same task many times a day for many days, how fast are likely to notice if something is going wrong, or if they will put their brains on autopilot and not notice.


One conference delegate suggested that the system could be used to help people prepare for underwater survival courses (required in preparation for going offshore), when people could see a visualisation of themselves getting out of a helicopter underwater before they are required to do it for real in a tank.

Another delegate suggested that it could be used to train people over longer scales, to see (for example) how long someone should be working on an offshore platform. Some people think that 90 days is a sensible limit. 'It's something that we'd love to take a look at,' Mr Prast replied.

One delegate suggested that it could be used to model 'citadels' on maritime vessels, safe rooms which the crew can go to if the vessel is under attack from pirates, and which weapons the rooms would be safe from.

Mr Prast said that he did a similar study in 2006 of what happens when soldiers weld metal bars themselves to their vehicles for additional protection. Sometimes this can actually increase the risk, if it means that a bullet ends up rattling around inside for longer, rather than being able to escape.

'We are not the experts in vehicle design and armour design - but we create the model which allows people who are the experts to test various designs,' he said.

One delegate, a safety advisor to the European Commission, said that there might be problems accessing data from the oil and gas industry, since it is 'not a particularly transparent industry.'

Mr Prast said that the process does not need a lot of data. 'If we have the procedures, if we know what people are doing or supposed to be doing, we don't need an exhaustive history of data for what has happened on the platform,' he said. 'We create a model that's based from bottom up.'

Associated Companies
» EOS Solutions

comments powered by Disqus


To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.


Latest Edition Jul-Aug 2022
Aug 2022

Download latest and back issues


Learn more about supporting Digital Energy Journal