π Digital π Industry π Health and biotech
Digital twins: what opportunities for industry?

Modelling the ocean enables advances in maritime tech

Anders Thorin, Research Engineer at the Interactive Simulation Laboratory of CEA List
On February 1st, 2023 |
4 min reading time
Anders Thorin
Research Engineer at the Interactive Simulation Laboratory of CEA List
Key takeaways
  • A digital twin is the digitisation of a given object or environment, which in the case of an interactive digital twin can be interacted with.
  • The CETO software uses this technology to address issues that arise in the maritime industry.
  • A digital twin allows scenarios to be validated before being implemented in the physical world.
  • The digital twin brings three major advantages: cost, safety, and reproducibility.
  • Digital twins not only allow scenarios to be predicted but can also be used to safely train personnel.

A dig­i­tal twin is the digi­ti­sa­tion of a giv­en envi­ron­ment: this very gen­er­al def­i­n­i­tion encom­pass­es many very var­ied appli­ca­tions. Inter­ac­tive dig­i­tal twins allow a human to inter­act with the vir­tu­al envi­ron­ment, thanks to track­ing sen­sors, con­trol organs (joy­stick, joy­sticks, etc.) or force feed­back inter­faces (hap­tic inter­faces). These tools are inter­dis­ci­pli­nary in sev­er­al respects: in terms of their fields of appli­ca­tion (med­i­cine, mar­itime indus­try, etc.) and the sci­ences they mobilise (mechan­i­cal, ther­mal, bio­log­i­cal, etc.). 

Anders Thorin, a research engi­neer at the CEA List’s Inter­ac­tive Sim­u­la­tion Lab­o­ra­to­ry, has been work­ing in the team devel­op­ing the XDE Physics soft­ware for the past twen­ty years. This soft­ware, which allows the cre­ation of dig­i­tal twins in the field of robot­ics, has attract­ed the inter­est of Tech­nip Ener­gies for projects in the mar­itime sec­tor. The researchers then devel­oped new func­tion­al­i­ties to meet the needs of the com­pa­ny. Today, it has been repack­aged in a soft­ware pack­age called “CETO”, ded­i­cat­ed to inter­ac­tive mar­itime sim­u­la­tion. It allows immer­sion in vir­tu­al real­i­ty (VR) with­in a crane on board a float­ing struc­ture, which can be used for fore­cast­ing and risk assess­ment for com­plex lift­ing oper­a­tions, or for staff training.

Replicating the sea and its environment

CETO makes it pos­si­ble to repro­duce the phys­i­cal char­ac­ter­is­tics of the sea, its envi­ron­ment, and numer­ous objects from the mar­itime world – con­tain­er ships, port cranes, cables, pipes – in an inter­ac­tive simulation. 

The first step was the phys­i­cal­i­sa­tion of the envi­ron­ment to be stud­ied – a step in which the researchers will mod­el the con­stituent ele­ments of the sce­nario, start­ing with the sea. “To mod­el the sea, we typ­i­cal­ly use 600 spec­tral com­po­nents,” says the researcher. Each spec­tral com­po­nent cor­re­sponds to a sinu­soidal wave, which has sev­er­al para­me­ters: ampli­tude (from the peak to the trough of the wave), direc­tion, phase, and fre­quen­cy. Oth­er envi­ron­men­tal con­di­tions must be con­sid­ered dur­ing the inter­ac­tive sim­u­la­tion: wind and cur­rent, for example.”

Super­po­si­tion of sine waves to gen­er­ate an irreg­u­lar sig­nal, source: Marinet1

After an ini­tial sim­u­la­tion pro­to­type, the researchers were able to improve the pre­ci­sion and there­fore the real­ism of the sim­u­la­tion by increas­ing the lev­el of detail of the mod­els: tak­ing into account pro­pellers, cables, cranes, etc., with a view to know­ing, or rather ver­i­fy­ing, how it will react to the move­ment of the sea, the wind, the swell and so on. “Phys­i­cal­i­sa­tion is not always nec­es­sary depend­ing on the pur­pose,” he says, “but when it is, the inter­ac­tive nature means that the phys­i­cal equa­tions have to be solved in real time, which often involves a sim­pli­fi­ca­tion phase of the phys­i­cal models.”

For exam­ple, a slen­der rigid struc­ture is mod­elled by a ‘beam’ of a cer­tain size, which is a sim­pli­fi­ca­tion to allow real-time sim­u­la­tion in XDE Physics. With this step, the ques­tion may arise: can a dig­i­tal twin be so accu­rate that it repro­duces real­i­ty, despite being a sim­pli­fi­ca­tion? The researcher no longer asks this ques­tion: “in sci­ence, every­thing is a mod­el. Even a con­cept as sim­ple as a right angle does not exist in nature. The chal­lenge for a dig­i­tal twin in inter­ac­tive sim­u­la­tion is to adopt a mod­el that is suf­fi­cient­ly pre­cise to be of prac­ti­cal use in a giv­en con­text. In addi­tion, the mod­el equa­tions cho­sen must be able to be solved quick­ly enough with the hard­ware pro­vid­ed. Per­fect accu­ra­cy is not required to obtain use­ful results, and so much the bet­ter, as this is not attainable.”

A good exam­ple of this is putting a deformable pipe in the water. Once the pipe has been mod­elled as a beam, the lab­o­ra­to­ry team will be able to assess its strength dur­ing lift­ing and low­er­ing oper­a­tions, depend­ing on the weath­er con­di­tions and the actions of the crane oper­a­tor in the vir­tu­al world. This mul­ti­tude of dig­i­tal­ly com­bined ele­ments allows for an inter­ac­tive sim­u­la­tion of any desired sce­nario. This allows the user to assess whether the pipe will with­stand han­dling with­out hav­ing to achieve per­fect accu­ra­cy in the sim­u­la­tion. All this is much safer and more eco­nom­i­cal than real-life testing. 

From forecasting to training

A dig­i­tal twin can there­fore allow sce­nar­ios to be val­i­dat­ed before being applied in the phys­i­cal world. The uses are there­fore count­less, and the advan­tages con­sid­er­able. “There is a real ben­e­fit to this type of sim­u­la­tion,” says Anders Thorin. “First­ly, in terms of cost, because it requires less time to car­ry out a project, lim­its the need to move equip­ment, avoids the use of mod­els, etc. And in terms of repro­ducibil­i­ty, because each sce­nario can be val­i­dat­ed on the same sim­u­la­tion, the para­me­ters being able to be mod­i­fied as required.”

The uses of the dig­i­tal twin are count­less, and the ben­e­fits considerable.

In fact, if we want to esti­mate the dif­fer­ence in move­ment that a ship can have in calm water, com­pared to rough water, only the para­me­ters that influ­ence the dynam­ics of the water and the wind need to be changed, name­ly: the direc­tion of the wind, its speed, and its con­stan­cy, etc. “It is a form of fore­cast­ing an event under cer­tain pre­de­fined con­di­tions – wind, swell, cur­rent, and many oth­ers,” adds the researcher.

How­ev­er, the use­ful­ness of this type of dig­i­tal twin does not stop at sce­nario ‘fore­cast­ing’. Hav­ing a real vir­tu­al world, acces­si­ble through a VR head­set, all as true as pos­si­ble to the phys­i­cal real­i­ties of the ele­ments around us, would allow for effec­tive use in train­ing. “Tech­nip Ener­gies adjust­ed a chair to repli­cate (this time in the real world) the cab­in of a crane with its joy­sticks,” he explains. “This allowed a crane oper­a­tor to come on site for train­ing, in addi­tion to hands-on practice.”

Cou­pled with its two main ben­e­fits (cost and repro­ducibil­i­ty), such soft­ware would have an unde­ni­able added val­ue to crane oper­a­tor train­ing, espe­cial­ly for deal­ing with extreme sit­u­a­tions, such as high winds.

Pablo Andres
1Marinet: Best prac­tice man­u­al for wave sim­u­la­tion – https://www.marinet2.eu/wp-content/uploads/2017/04/D2.8‑Best-Practice-Manual-for-Wave-Simulation.pdf

Our world explained with science. Every week, in your inbox.

Get the newsletter