Physical AI da kula da ƙwadago: amana, alhaki da sake fasalin aiki

TL;DR. A taron Physical AI and Robotics Summit a London, Kier Group da BSI sun tattauna amanar ma'aikata, tsaron mika iko ga mutum, da raba alhaki a masana'antu.

Mahimman Bayanai

Girgizar jiki na zahiri: fiye da automation na kan allo. A zaman kwamitin A meeting of minds (and machines): Humans, robots and the redesign of work (Haɗuwar tunani da injuna: mutane, robot da sake fasalin aiki), shugabannin kamfanoni sun yi nazarin rikicin tunani na…
Matsalar mika iko: magance “dark side autonomy”. Wata babbar muhawarar fasaha a zaman ta mai da hankali kan raunin tsaro mai tsanani da ke cikin cin gashin kai na rabi, batun da masana robotics da injiniyoyin human factors ke kira “dark side autonomy” (ɓangaren…
Kwatanta tsare-tsare: samfuran shugabancin aiki a robotics na masana'antu. Tushe: Haɗakar gudummawar kwamitin, binciken occupational psychology da ƙa'idojin shugabancin haɗari na kayayyakin more rayuwa da aka gabatar a taron Economist Enterprise Physical AI and Robotics Summit (Convene 200…
Abin da wannan ke nufi a kowane fanni. Ana aiwatar da sake fasalin aiki ta fuskar ayyuka, tsaro da shari'a ta hanyoyi daban-daban a fannonin masana'antu na zahiri.

Idan software na kamfani ya lalace, yana ɓata bayanai a rumbun adana bayanai, yana katse haɗin hanyar sadarwa, ko kuma yana fallasa wani token. Idan tsarin physical AI ya lalace, ɗaruruwan kilogiram na ƙarfe mai gaɓoɓi, masu tura ruwa (hydraulic actuators) masu ƙarfin murɗawa, ko injunan haƙa ƙasa masu cin gashin kansu suna karo da kayayyakin more rayuwa ko jikin mutane. A taron Economist Enterprise Physical AI and Robotics Summit da aka gudanar a Convene 200 Aldersgate a London a ranar 13 ga Oktoba 2026, wani babban kwamiti ya taru don tattauna iyakokin aiki, ɗa'a da shari'a na cin gashin kan injuna a masana'antu: yadda za a kiyaye amanar ma'aikata, a rage kasawar mika iko ta fuskar tunani, da kuma kafa tsarin alhaki wanda doka za ta iya karewa yayin da injuna masu basira ke shiga sahun gaba.

Girgizar jiki na zahiri: fiye da automation na kan allo #

A zaman kwamitin A meeting of minds (and machines): Humans, robots and the redesign of work (Haɗuwar tunani da injuna: mutane, robot da sake fasalin aiki), shugabannin kamfanoni sun yi nazarin rikicin tunani na musamman da ke tasowa idan artificial intelligence ya bar allo ya shiga wurin aiki na zahiri.

Laura Bishop, AI & Cyber Security Digital Sector Lead a British Standards Institution (BSI) kuma ƙwararriya a ilimin halin ɗan adam a wurin aiki (occupational psychology), ta bayyana babban ƙalubalen: idan ma'aikata suna amfani da manyan language model ko copilot na kwamfuta, suna kiyaye nisa ta fuskar motsin rai. Software yana cikin tashar gilashi; ana iya rage shi ko a kashe shi.

Sabanin haka, physical AI yana mamaye sarari na zahiri. Excavator mai cin gashin kansa da ke aiki a wani babban aikin babbar hanya, hannun robot na haɗin gwiwa mai motsi da ke ɗaukar akwatuna tare da ma'aikatan rumbu, ko na'urar rushe gini ta robot da ke motsawa cikin ginin kasuwanci mai ƙunci, duk suna raba muhallin zahiri ɗaya da ma'aikata mutane. Idan ma'aikata suka ga inji mai ƙarfin jiki da kuma ikon yanke shawara da kansa, adawar tunani tana ƙaruwa. Ma'aikata suna fuskantar abin da Bishop ta kira “barazana biyu”: tsoron rasa rayuwa ta fuskar tattalin arziki tare da damuwa nan take game da tsaron jiki.

Louisa Finlay, Chief Operating Officer na Kier Group (ɗaya daga cikin manyan ƙungiyoyin kayayyakin more rayuwa, gine-gine da ayyuka a United Kingdom), ta tabbatar da cewa ba za a iya tilasta nasarar karɓar fasaha a manyan masana'antu ta umarni daga sama ba. Ma'aikatan sahun gaba a gine-gine, aikin ƙasa da kayayyakin more rayuwa mutane ne masu ra'ayin aiki a zahiri. Idan aka ɗauki wata fasaha a matsayin na'urar leƙen asiri, haɗari ga tsaro, ko hanyar rage yawan ma'aikata, ma'aikata za su yi adawa da ita, su kauce mata, ko su lalata ta.

Hanyar Kier Group ta ta'allaka ne kan hulɗa kai tsaye, ta gani kuma a bayyane. Finlay ta jaddada cewa dole ne shugabanni su kasance masu gaskiya ƙwarai game da sauye-sauyen ƙungiya. Ayyukan gudanarwa da na bin diddigi da hannu za su ragu; sai dai injuna za su iya kawar da ayyukan da ke jawo raunin tsoka da ƙashi na dogon lokaci, silicosis, da hatsarorin masana'antu masu kisa.

Matsalar mika iko: magance “dark side autonomy” #

Wata babbar muhawarar fasaha a zaman ta mai da hankali kan raunin tsaro mai tsanani da ke cikin cin gashin kai na rabi, batun da masana robotics da injiniyoyin human factors ke kira “dark side autonomy” (ɓangaren duhu na cin gashin kai).

A tsarin masu cin gashin kansu na Level 2 da Level 3 (ko motoci masu tuƙa kansu ne ko injunan masana'antu na haɗin gwiwa), robot yana gudanar da ayyuka da kansa a yanayi na yau da kullum, amma yana buƙatar mai sarrafawa ɗan adam ya sa ido kan tsarin kuma ya karɓi iko da hannu duk lokacin da wani abu mara kyau ya taso.

Laura Bishop da Roman, mai tura tsare-tsare, sun yi gargaɗin cewa wannan tsarin kulawa yana da kuskure mai haɗari ta fuskar tunanin ɗan adam:

Ɗauke nauyin tunani da raguwar faɗakarwa: Kwakwalwar ɗan adam ba ta iya ci gaba da faɗakarwa sosai yayin da take kallon inji yana yin ayyuka iri ɗaya marasa kuskure na sa'o'i. Cikin minti ashirin, hankalin mai sarrafawa yana karkacewa, lokacin mayar da martani yana ninkuwa, kuma bin diddigin ido yana tabbatar da janyewar tunani.
Gibin dawo da fahimtar halin da ake ciki: Idan wani lamari na gefe da ba a zata ba ya faru (kamar layin kayan aiki na ƙarƙashin ƙasa da ba a taswira ba da aka ci karo da shi yayin da robot ke haƙa rami, ko mai tafiya a ƙafa da ya shiga wurin da ba a gani), tsarin mai cin gashin kansa yana kunna ƙararrawar gaggawa ba zato ba tsammani kuma yana mayar da iko ga mai sarrafawa. Bincike ya nuna cewa dawo da cikakkiyar fahimtar halin da ake ciki ta gani da sarari, fahimtar yanayin lalacewar, da aiwatar da tsoma baki na zahiri mai tasiri yana buƙatar daƙiƙa biyar zuwa goma sha biyar. A ayyukan motsi masu sauri, wannan jinkiri yana kashe rai.
Moral crumple zone: Dogaro da kulawar ɗan adam mara aiki sau da yawa yana zama garkuwar kamfani daga alhaki, yana mayar da ma'aikaci ɗan adam “moral crumple zone” (yankin ɗaukar laifi): shi ne ke ɗaukar laifin shari'a da na laifi kan kasawar algorithm ta tsarin mai sarrafa kansa saboda kawai yana da ikon soke umarni a suna.

Matsayar da BSI da Kier Group suka cimma ita ce dole ne tsarin masana'antu su koma ga yankunan tsarin aiki da ake sa ido a kansu (monitored operational design domains, ODD) masu iyakokin tsayawa na fail-operational ko fail-safe. Idan inji mai cin gashin kansa ya ci karo da wani abu mara kyau a wajen iyakokin da aka tabbatar masa, kada ya mika iko ba zato ba tsammani ga mutum da bai shirya ba; dole ne ya koma da kansa zuwa yanayin tsaro mai hasashe ba tare da kuzari ba (kamar sanya birki, sauke boom da yanke matsin ruwan hydraulic) yayin da yake neman mutum ya duba.

Kwatanta tsare-tsare: samfuran shugabancin aiki a robotics na masana'antu #

Ginshiƙin shugabanci	Samfurin masu fara amfani ba tare da ƙa'ida ba	Mai kulawa mara aiki (“mutum a matsayin mai tsaron baya”)	Tsarin raba alhaki da aka tabbatar (ƙa'idar BSI / Kier)
Sadarwa da ma'aikata	Ba a bayyane ba; ana kawo automation ba zato ba tsammani don rage farashin kowace naúra nan take	Horon bin ƙa'ida daga sama; ana gaya wa ma'aikata su “kalli injin”	Tattaunawa a bayyane; ma'aikata suna shiga ƙirar kayan aiki, sabon horo da binciken tsaro
Ka'idar mika iko ta tunani	Ƙararrawa ba zato ba tsammani da ke buƙatar tsoma bakin mutum nan take	Ana sa ran soke umarni da hannu cikin daƙiƙa 1 zuwa 2 bayan matsalar tsarin	Yankunan aiki masu iyaka; komawa kai tsaye zuwa tsayawa fail-safe ba tare da kuzari ba
Raba alhaki	Ba a ayyana ba; an binne shi cikin sanarwar ƙin alhaki ta mai sayarwa	Ana tura shi ƙasa kan mai sarrafawa na sahun gaba ko mai kula da wurin aiki	Matakin kwangila mai matakai uku (Model Provider, Systems Integrator, Site Deployer)
Bin ƙa'idojin doka	Kaɗan; tabbatar da kai na asali a ƙarƙashin machinery directive	Shingen robot na masana'antu na yau da kullum bisa ISO 10218-1/2	Cikakken bin ISO/TS 15066, ƙa'idojin BSI PAS da umarnin EU AI Act na High-Risk
Rubuta inshora	Kuɗin inshora masu tsada; kariya mai sharaɗi saboda haɗarurrukan da ba a taswira ba	Alhakin kasuwanci na yau da kullum tare da sharuɗɗan subrogation masu tsauri	Rubuta inshora da aka tsara bisa telemetry da aka bincika da rajistar telemetry na lokacin aiki
Ayyana matsayin ma'aikaci	Ƙwadagon da aka maye gurbinsa ko mai kula da inji mai ƙaramin albashi	Mai sa ido mara aiki mai yawan matsin lamba da saurin gajiyar tunani	Mai kula da rukunin injuna da aka tabbatar, manajan teleoperation da jagoran dabaru na dijital

Tushe: Haɗakar gudummawar kwamitin, binciken occupational psychology da ƙa'idojin shugabancin haɗari na kayayyakin more rayuwa da aka gabatar a taron Economist Enterprise Physical AI and Robotics Summit (Convene 200 Aldersgate, London, 13 ga Oktoba 2026).

Abin da wannan ke nufi a kowane fanni #

Ana aiwatar da sake fasalin aiki ta fuskar ayyuka, tsaro da shari'a ta hanyoyi daban-daban a fannonin masana'antu na zahiri.

Manyan gine-gine da kayayyakin more rayuwa #

A kayayyakin more rayuwa na farar hula, ƙarancin ma'aikata, tsufan al'umma da haɗarurrukan tsaro masu tsanani suna haifar da buƙatar automation cikin gaggawa. Gine-gine na ci gaba da kasancewa ɗaya daga cikin masana'antu mafi haɗari a duniya, wanda ke jawo dubban munanan raunuka da mutuwa kowace shekara daga faɗuwa daga tsayi, rushewar gine-gine da karo da manyan injuna.

Louisa Finlay ta bayyana yadda Kier Group ke sake fasalin ayyukan sahun gaba ta hanyar sarrafa aiki mai haɗari na zahiri da injuna. Tura turaku ta robot mai cin gashin kansa, injunan matse ƙasa masu sarrafa kansu, da auna girman ƙasa ta jirgi mara matuƙi suna cire ma'aikata mutane daga gefen ramuka masu haɗari da hanyoyin jigilar kaya masu aiki.

Abu mafi muhimmanci shi ne masu sarrafa injuna da a da suke kwashe sa'o'i takwas cikin ɗakunan diesel masu girgiza, suna fuskantar ƙura da hayaniya, yanzu ana horar da su su zama manajojin wurin aiki na dijital. Daga ɗakunan sarrafa wurin aiki masu na'urar sanyaya iska ko cibiyoyin ayyuka na nesa, waɗannan ma'aikata suna kula da injuna masu cin gashin kansu na rabi da yawa a lokaci guda, ta amfani da bidiyo masu inganci da telemetry na digital twin. Gamsuwar ma'aikata tana ƙaruwa domin gajiyar jiki da haɗarin rauni suna raguwa yayin da albashi ke daidaita da matakan ƙwarewar fasaha mafi girma.

Ayyukan jama'a da gyaran kayayyakin farar hula #

A tace ruwa, rushe tashoshin nukiliya da gyaran layukan wutar lantarki, ma'aikata suna fuskantar muhallai masu guba, masu yawan radiation ko masu ƙarfin wuta mai yawa. Robot masu rarrafe na bincike da jirage marasa matuƙi masu cin gashin kansu suna maye gurbin shigar mutane cikin magudanan ruwa masu ƙunci da tashoshin rarraba wuta.

Shugabanci a wannan fanni yana mai da hankali sosai kan ingancin cyber-physical. Kamar yadda Laura Bishop ta lura, tsarin physical AI na'urorin IoT ne masu haɗi da ke da masu tura motsi. Idan aka kutsa cikin tsaron yanar gizo na robot mai cin gashin kansa da ke auna sinadarai ko na'urar sarrafa bawul na bututu, sakamakon ba satar bayanai ba ne; gurɓata muhalli ne na zahiri. Shugabancin aiki yana buƙatar rajistar bincike da ba za a iya canzawa ba wadda aka sa hannu ta hanyar cryptography, da raba kayan aiki ta zahiri (air-gap) a wuraren katse wuta masu muhimmanci ga tsaro.

Dabaru da hanyoyin cika oda #

Cibiyoyin cika oda na e-commerce sun kasance babban filin gwaji ga robot masu motsi na haɗin gwiwa masu cin gashin kansu (AMR). Sai dai aiwatarwar farko da ta ɗauki ma'aikata a matsayin masu ɗauko kaya cikin sauri da algorithm na na'urar ɗauka ta robot ke tafiyar da su ta jawo yawan barin aiki da rikici da ƙungiyoyin ma'aikata.

Shugabancin dabaru na zamani yana fifita haɗin gwiwa mai la'akari da jin daɗin jiki. Kekunan ɗauka masu cin gashin kansu suna ɗaukar manyan akwatuna tsakanin hanyoyin ɗauko kaya, suna barin ma'aikata mutane a wuraren ɗauko kaya da aka tsara don jin daɗin jiki inda suke bincika kaya ta taɓawa da rarraba su. Ma'aunin tsarin yana sauyawa daga adadin ɗauko kaya na kowane minti mai hukunci zuwa daidaiton cika oda na ƙungiya baki ɗaya, yana kafa al'adar dogaro da juna tsakanin mutum da inji.

Injiniyancin kwangila: tsara alhaki mai matakai da yawa #

Wa ke da alhakin shari'a idan robot mai cin gashin kansa ya jawo lalacewa ta zahiri ko rauni? Kwamitin ya tattauna wannan tambaya kai tsaye, ya ƙi ra'ayin mai sauƙi cewa alhaki zai iya kasancewa kan ɓangare ɗaya kaɗai.

A physical AI na zamani, tsarin masu cin gashin kansu haɗaka ce mai sarƙaƙiya ta foundation model na asali, fine-tuning na musamman, na'urorin ji na kasuwanci, jikin inji da haɗawa a matakin wurin aiki.

Kamfanonin da ke tura tsarin suna kafa kwangilolin raba haɗari masu matakai uku:

Mataki na 1: masu samar da foundation model da kayan aikin OEM #

Mai ƙera kayan asali (OEM) da mai haɓaka foundation model suna ɗaukar alhakin lahani na asali a cikin algorithm, lalacewar kayan aiki ta inji, da keta iyakokin tsaro da aka horar a baya. Idan wani model ya nuna hallucination da ke tsallake iyakokin tsaro na asali a kan bayanan shigarwa da ke cikin bayanin da aka tabbatar masa, alhakin yana kan mai samarwa.

Mataki na 2: masu haɗa tsarin da masu tura shi a fanni #

Mai haɗa tsarin da ke da alhakin fine-tuning na model kan bayanai na musamman na fanni, daidaita kyamarori da na'urorin LiDAR, da tsara ka'idojin iyaka na gida yana ɗaukar alhakin lahanin haɗawa. Dole ne kwangilolin masu haɗawa su ayyana Environmental Operational Design Domain (ODD) a fili, suna bayyana ma'aunin haske, yanayi, matakin ƙura da gangaren ƙasa da aka yarda da su.

Mataki na 3: kamfanonin da ke gudanar da aiki da masu wurin aiki #

Kamfanin da ke tura robot a wurin aiki yana da alhakin bin ƙa'idojin muhalli, kula da wurin aiki na zahiri da horar da ma'aikata. Idan manajan wurin aiki ya bar mutane marasa izini su shiga yankin aikin robot mai aiki, ya kashe geofence na tsaro na zahiri, ko ya kasa bin jadawalin daidaitawa da aka wajabta, alhakin yana kan mai gudanar da wurin aiki.

Ta hanyar shigar da waɗannan iyakoki cikin kwangilolin kasuwanci da daidaita su da buƙatun rubuta inshora, kamfanoni za su iya tura physical AI da tabbaci na shari'a a bayyane maimakon jiran hukuncin kotu bayan bala'in masana'antu.

Kammalawa #

Fahimtar da Louisa Finlay, Laura Bishop da sauran mahalarta kwamitin suka raba a taron Economist Enterprise Physical AI and Robotics Summit ta bayyana gaskiya ɗaya a fili: babbar matsalar faɗaɗa physical AI a ayyukan kamfanoni ba ƙwarewar neural networks ba ce ko ƙarfin brushless actuators; shugabancin ɗan adam ne.

Robot suna aiki a duniyar zahiri, tare da mutane waɗanda rayuwarsu, mutuncinsu da tsaron jikinsu suka dangana da muhallin aikinsu. Duk wata ƙungiyar shugabancin kamfani da ta ɗauki physical AI a matsayin aikin kuɗi zalla na kawar da farashin ƙwadago za ta fuskanci tsananin adawar al'ada, munanan kasawar tsaro wajen mika iko, da alhakin doka mai nauyi.

Haƙiƙanin fifikon aiki na ƙungiyoyin da ke ɗaukar physical AI a matsayin mai haddasa sake fasalin aiki mai zurfi ne. Ta hanyar kawar da aiki mai haɗari, girmama iyakokin tunani yayin hulɗar mutum da inji, da tsara layukan alhaki na kwangila a fili, kamfanoni masu hangen nesa za su iya gina makoma inda injuna masu basira da ma'aikata mutane masu ƙarfi ke haɗa kai don gina masana'antu masu tsaro da juriya.

Tambayoyin da Ake Yawan Yi #

Me ya sa physical AI ke tayar da damuwar ma'aikata fiye da AI na software? Sabanin automation na software (kamar chatbot ko predictive analytics) wanda ke kan allo kawai, physical AI yana aiki a duniyar zahiri, yana amfani da kuzarin motsi, nauyi da sauri. Ma'aikata suna fuskantar “barazana biyu” mai tsanani: tsoron rasa aiki ta fuskar tattalin arziki tare da damuwa mai zurfi game da tsaron jikinsu a wurin aiki da ake rabawa.

Mene ne “dark side autonomy” a robotics? “Dark side autonomy” na nufin kasawar tunani da ke faruwa idan aka sanya masu sarrafawa mutane su sa ido ba tare da aiki ba kan tsarin masu cin gashin kansu na rabi a matsayin masu tsaron baya. Mutane ba su iya ci gaba da faɗakarwa a dogon lokaci na sa ido mara aiki. Idan matsala ta gefe ta faru, masu sarrafawa suna fama da ƙarancin fahimtar halin da ake ciki kuma ba za su iya tsoma baki da sauri don hana hatsari ba.

Ta yaya Kier Group ke tunkarar karɓar physical AI a gine-gine? Kier Group tana mai da hankali kan hulɗa a bayyane kuma ta gani da ma'aikatan sahun gaba. Maimakon gabatar da robotics a matsayin kayan rage ma'aikata, Kier tana gabatar da automation a matsayin hanyar kawar da aiki mai haɗari da mai raunana jiki - kamar haƙa ramuka, aiki a tsayi da ɗaga kaya masu nauyi. Ana sake horar da ma'aikata zuwa matsayin kulawa, suna sarrafa injuna masu cin gashin kansu na rabi da yawa daga nesa a muhallai masu tsaro da aka sarrafa.

Yaya ya kamata a raba alhakin kwangila a hatsarorin robotics masu cin gashin kansu? Ya kamata a raba alhaki a matakai uku bisa ikon yanke shawara: masu samar da foundation model da OEM suna da alhakin kasawar algorithm a cikin ma'aunin da aka tabbatar; masu haɗa tsarin suna da alhakin daidaitawa na musamman ga wurin aiki, ɗora na'urorin ji da daidaitawa ga fanni; kuma masu tura tsarin a wurin aiki suna da alhakin kiyaye geofencing na wurin aiki, tsaron muhalli da horar da masu sarrafawa.

Waɗanne ƙa'idoji ne ke jagorantar tura robot na masana'antu na haɗin gwiwa cikin tsaro? Manyan ƙa'idojin sun haɗa da ISO 10218-1 da ISO 10218-2 (buƙatun tsaro ga robot na masana'antu da haɗa tsarin), ISO/TS 15066 (takamaiman buƙatun tsaro ga ayyukan robot na haɗin gwiwa da iyakokin ƙarfin taɓawar mutum da robot), sabbin jagororin British Standards Institution (BSI) kan shugabancin tsarin masu cin gashin kansu, da buƙatun daidaito na EU AI Act na High-Risk Machinery.

Manazarta #

Economist Enterprise, (2026). Physical AI & Robotics Summit: fifikon aiki ta hanyar tsare-tsare masu cin gashin kansu ⧉. Economist Enterprise.
Finlay, L., (2026). A meeting of minds (and machines): mutane, robot da sake fasalin aiki ⧉. Kier Group.
Bishop, L., (2026). Human factors, tsaron tunani da shugabanci a tsarin masu cin gashin kansu ⧉. British Standards Institution (BSI).
British Standards Institution, (2025). BSI PAS 2025: shugabanci da tura tsarin masu motsi masu cin gashin kansu bisa ɗa'a a muhallan masana'antu ⧉. BSI.
European Parliament and Council, (2024). Regulation (EU) 2024/1689 mai shimfiɗa ƙa'idoji ɗaya kan artificial intelligence (Artificial Intelligence Act) ⧉. Official Journal of the European Union.
Bainbridge, L., (1983). Ironies of Automation: abubuwan mamaki na automation ⧉. Automatica, 19(6), 775-779.
International Organization for Standardization, (2026). ISO/TS 15066:2026 Robot da na'urorin robot - Robot na haɗin gwiwa ⧉. ISO.

Bita ta ƙarshe 2026-10-13.

Bita ta ƙarshe 2026-10-27.

Sake buga wannan labari

Physical AI da kula da ƙwadago: amana, alhaki da sake fasalin aiki — Sebastien Rousseau

A taron Physical AI and Robotics Summit a London, Kier Group da BSI sun tattauna amanar ma'aikata, tsaron mika iko ga mutum, da raba alhaki a masana'antu.

An lasisin wannan labari a karkashin Creative Commons Attribution 4.0 International. Sake bugawa na bukatar nuna asalin URL na asali.

Physical AI da kula da ƙwadago: amana, alhaki da sake fasalin aiki — Sebastien Rousseau

A taron Physical AI and Robotics Summit a London, Kier Group da BSI sun tattauna amanar ma'aikata, tsaron mika iko ga mutum, da raba alhaki a masana'antu.

Originally published at https://sebastienrousseau.com/ha/2026-10-13-physical-ai-kula-kwadago-amana-alhaki/ by Sebastien Rousseau.
Licensed under CC-BY-4.0.

SEBASTIEN ROUSSEAU FOUNDER · ENGINEER