Thought Leadership article from Ansys: How simulation can support safety in advanced air mobility

Thought lead­er­ship arti­cle by Bernard Dion, Fel­low and Chief Tech­nol­o­gist for Sys­tems and Embed­ded Soft­ware at Ansys.

Across the avi­a­tion indus­try, advanced air mobil­i­ty (AAM) is becom­ing a major focus for orig­i­nal equip­ment man­u­fac­tur­ers (OEMs) and sup­pli­ers alike.

Com­pa­nies are devel­op­ing air­craft for peo­ple and car­go that will aid under­served local, region­al, urban, and rur­al envi­ron­ments. And to make these inno­va­tions as effi­cient and effec­tive as pos­si­ble, auton­o­my has become a key focus.

This next gen­er­a­tion of air­craft will be small­er, and will fly at low­er alti­tudes. Whilst some are being based on exist­ing tech­nolo­gies, such as heli­copters, a com­pet­i­tive mar­ket for elec­tric ver­ti­cal take- off and land­ing (eVTOL) air­craft has emerged. Research around AAM in the UK esti­mates that up to 11.4 mil­lion jour­neys will be sub­sti­tut­ed by eVTOLs by 2040, result­ing in up to £2.1bn in annu­al socioe­co­nom­ic ben­e­fit. 

The wide­spread adop­tion of autonomous AAM has the poten­tial to rev­o­lu­tionise trans­port, and the indus­try, as we know it. Con­sumer demand already exists, with 57 per cent of peo­ple across the globe open to trav­el­ling in eVTOLs.

Yet, there is still some way to go – get­ting an air­craft in the air safe­ly is the high­est pri­or­i­ty, and reg­u­la­tion author­i­ties have strict stan­dards that com­pa­nies must meet. As autonomous tech­nolo­gies devel­op, uphold­ing safe­ty stan­dards becomes even more cru­cial. To adhere to reg­u­la­tions and decrease time to mar­ket, engi­neers can estab­lish a dig­i­tal thread, pow­ered by sim­u­la­tion. 

Address­ing chal­lenges around auton­o­my in AAM

Auton­o­my is based on three func­tions. The first is per­cep­tion based on sen­sors like cam­eras, lidars or radars. The sec­ond is deci­sion-mak­ing to avoid a col­li­sion by detect­ing an obsta­cle, for exam­ple. The third is actu­a­tion, which uses flight and engine con­trols to per­form the cor­rect manoeu­vre with­out human inter­ven­tion.

Devel­op­ing these autonomous func­tions, and ensur­ing their safe­ty, is an incred­i­bly com­plex and intri­cate process. Engi­neers must per­form a safe­ty analy­sis to analyse the oper­a­tional design domain (ODD), and out­line the chal­lenges faced dur­ing devel­op­ment.

A mul­ti­tude of fac­tors, such as the weath­er and elec­tron­ics fail­ure con­di­tions, must be tak­en into con­sid­er­a­tion to ensure that AAM air­craft can oper­ate effec­tive­ly. For exam­ple, extreme weath­er con­di­tions like fog can impede sen­sors and risk poor per­cep­tion and deci­sion-mak­ing. Elec­tron­ic fail­ures could also occur, which can impact the actu­a­tion func­tion and result in the wrong manoeu­vres.

This analy­sis is the most crit­i­cal and dif­fi­cult fac­tor in the devel­op­ment process, so engi­neers can turn to sim­u­la­tion to over­come poten­tial safe­ty chal­lenges. Sim­u­la­tion can be used to dig­i­tal­ly repli­cate and test any com­po­nent of an air­craft such as the pow­er­train or aero­dy­nam­ics. This pro­vides a clear, 360-degree view of how the air­craft will oper­ate in dif­fer­ent envi­ron­ments.

Addi­tion­al­ly, engi­neers can eas­i­ly cre­ate an end­less num­ber of poten­tial sce­nar­ios that a vehi­cle may encounter, allow­ing them to analyse how it would react, and then opti­mis­ing accord­ing­ly to guar­an­tee safe­ty. By dig­i­tal­ly repli­cat­ing the test­ing process, engi­neers can eas­i­ly pin­point and rec­ti­fy poten­tial errors to cre­ate a thor­ough safe­ty analy­sis to meet reg­u­la­tion author­i­ties’ stan­dards, and decrease the time to mar­ket.

Advanc­ing safe­ty in auton­o­my with dig­i­tal engi­neer­ing

Sim­u­la­tion is piv­otal for devel­op­ing and safe­ty test­ing auton­o­my in AAM, but devel­op­ing autonomous func­tions requires inter­con­nect­ed­ness. Engi­neers must there­fore incor­po­rate sim­u­la­tion into an over­all dig­i­tal envi­ron­ment to achieve break­throughs in auton­o­my and keep up with emerg­ing safe­ty reg­u­la­tions. With dig­i­tal engi­neer­ing, engi­neers can con­nect sim­u­la­tion to the process­es, tools, and data across prod­uct devel­op­ment, and can ensure safe­ty in com­plex prod­ucts.

This way, they can eas­i­ly col­lab­o­rate to vir­tu­al­ly design and test a prod­uct.  Cre­at­ing a dig­i­tal envi­ron­ment estab­lish­es a dig­i­tal thread, which enables trace­abil­i­ty and effec­tive col­lab­o­ra­tion between the AAM oper­a­tor and the air­craft devel­op­er.

This is vital to max­imis­ing pas­sen­ger safe­ty and meet­ing reg­u­la­tions, as dig­i­tal engi­neer­ing impacts more than just the devel­op­ment peri­od. A dig­i­tal thread allows engi­neers to mon­i­tor and update the prod­uct to ensure it oper­ates cor­rect­ly in the event of an error. As AAM evolves, teams can also eas­i­ly ensure that the prod­uct con­tin­u­al­ly meets devel­op­ing safe­ty stan­dards.

The role of arti­fi­cial intel­li­gence

As arti­fi­cial intel­li­gence and machine (AI and ML) become increas­ing­ly inte­grat­ed into auton­o­my func­tions, digi­tial engi­neer­ing can also help train the tech­nol­o­gy to max­imise safe­ty.

AI and ML are being lever­aged to devel­op the per­cep­tion func­tion in place of com­plex image pro­cess­ing to per­ceive the aircraft’s envi­ron­ment. They’re also being used in place of com­plex con­trol laws to improve deci­sion-mak­ing in uncer­tain envi­ron­ments.  

This added lay­er of intel­li­gence will accel­er­ate auton­o­my in AAM, as long as it’s devel­oped in an inter­con­nect­ed dig­i­tal envi­ron­ment. With dig­i­tal engi­neer­ing, sim­u­la­tion can be lever­aged to per­form super­vised learn­ing by pro­vid­ing images and solu­tions to train AI and ML so that they can devel­op the per­cep­tion func­tion.

Sim­u­la­tion can also train AI and ML in the deci­sion-mak­ing func­tion through rein­force­ment learn­ing, a tri­al-and-error process. Engi­neers can then see how this process will impact the over­all safe­ty of the air­craft, bring­ing them one step clos­er to the wide­spread adop­tion of AAM.

The future of AAM

The avi­a­tion indus­try is on the precipice of trans­for­ma­tion. Com­pa­nies are work­ing togeth­er to build the future of flight – but safe­ty is para­mount to the devel­op­ment of these AAM air­crafts. By lever­ag­ing sim­u­la­tion and estab­lish­ing a dig­i­tal thread, engi­neers can ensure that these inno­v­a­tive new tech­nolo­gies meet evolv­ing stan­dards and reg­u­la­tions, so that the future of AAM is as safe as pos­si­ble.

You can lis­ten to more of Bernard’s insights by lis­ten­ing to our lat­est pod­cast, which was released last week and is avail­able on whichev­er pod­cast plat­form you use.