Tsarin Haɗin Kai na Gano Matsayi Tsakanin Robobi da Wayoyin Hannu na Zamani Dangane da Sadarwar Haske da ake iya Gani

1. Bayyani Gabaɗaya

Wannan takarda tana magance kalubalen gano matsayi a cikin gida inda tsarin gargajiya kamar GPS ya gaza saboda toshewar siginar. Ta yi amfani da yaduwar fitilun LED da manyan na'urori masu daukar hoto (CMOS sensors) a cikin wayoyin hannu da robobi. Tsarin da aka gabatar yana amfani da Gano Matsayi ta Hanyar Haske da ake iya Gani (VLP), inda masu watsa LED ke daidaita haskensu (ta amfani da Maɓalli na Kashe-Kunna - OOK) don saka bayanan gani na musamman (UID) da bayanan matsayi. Tashar karɓa (kyamarar wayar hannu ko na'urar gani ta robot) tana ɗaukar waɗannan sauye-sauyen haske masu sauri ta hanyar tasirin rufe tagogi a jere (rolling shutter effect), wani abu da aka rubuta sosai a cikin binciken sadarwar kyamara ta gani (OCC). Wannan yana ba da damar watsa bayanai fiye da ƙimar firam ɗin bidiyo. Ta hanyar fassara tsarin hasken da aka ɗauka ("tsallake-tsallake") don dawo da UID kuma a yi nazari tare da ma'ajin taswira da aka adana a baya, na'urar za ta iya tantance matsayinta daidai. Takardar ta sanya wannan fasaha a matsayin mai ba da damar haɗin gwiwar mutum da robot a cikin yanayi mai motsi kamar ɗakunan ajiya da ayyukan kasuwanci, inda sanin yanayi na gaskiya da raba lokaci-lokaci ya fi muhimmanci.

2. Ƙirƙira

Babban ƙirƙira yana cikin tsarin haɗin kai kanta. Yayin da aka bincika VLP don na'urori masu zaman kansu, wannan aikin ya haɗa gano matsayi na wayoyin hannu da robobi cikin tsarin haɗin kai ɗaya. Manyan gudunmawar sun haɗa da:

1. Ƙirar Tsarin Aiki: Tsarin gano matsayi na haɗin kai dangane da VLC wanda aka keɓance don kalubalen amfani da wayar hannu (misali, karkatar da na'ura) da kewayawar robot, yana amfani da tsare-tsare na VLP da yawa don ƙarfi.
2. Aiwar Tsarin Aiki: Tsarin aiki inda ake samun matsayin robobi da wayoyin hannu kuma ake raba su cikin sauri, ana ganin su akan fuskar wayar hannu.
3. Tabbatar da Gwaji: Mai da hankali kan tabbatar da daidaiton gano ID, daidaiton gano matsayi, da aikin gaskiya-lokaci.

3. Bayanin Nunin Gwaji

Tsarin nunin gwaji ya kasu kashi biyu zuwa masu watsawa da masu karɓa.

4. Cikakkun Bayanai na Fasaha & Tsarin Lissafi

Tsarin ya dogara ne akan ƙa'idodin asali na OCC da gano matsayi ta hanyar lissafi.

1. Daidaitawar OOK & Tasirin Rufe Tagogi a Jere:
LED yana watsa jerin binary. Ana wakiltar '1' da LED KUNNA, kuma '0' da KASHE (ko akasin haka). Rufe tagogi a jere na kyamarar wayar hannu yana fallasa layuka daban-daban na na'urar firikwensin a lokuta daban-daban kaɗan. Lokacin ɗaukar LED mai ƙyalli cikin sauri, wannan yana haifar da ƙungiyoyi masu haske da duhu a cikin hoton. Tsarin waɗannan ƙungiyoyin yayi daidai da jerin bit ɗin da aka watsa. Ƙimar bayanai $R_{data}$ tana iyakance ta ƙimar samfurin rufe tagogi a jere, ba ƙimar firam ɗin $FPS$ ba: $R_{data} \approx N_{rows} \times F_{rs}$, inda $N_{rows}$ shine adadin layukan firikwensin kuma $F_{rs}$ shine mitar duban layi.

2. Ƙididdigar Matsayi:
Da zarar an dawo da matsayin 3D na LED $n$ daga ma'ajin ($\mathbf{P}_{LED,i} = [x_i, y_i, z_i]^T$), kuma an samo abubuwan da suka dace na 2D a kan farantin hoto ($\mathbf{p}_i = [u_i, v_i]^T$), ana iya ƙididdige matsayin 6-DOF na kyamara (matsayi $\mathbf{t}$ da alkibla $\mathbf{R}$) ta hanyar warware matsalar Perspective-n-Point (PnP): $$ s_i \begin{bmatrix} u_i \\ v_i \\ 1 \end{bmatrix} = \mathbf{K} [\mathbf{R} | \mathbf{t}] \begin{bmatrix} x_i \\ y_i \\ z_i \\ 1 \end{bmatrix} $$ inda $s_i$ shine ma'aunin sikelin, kuma $\mathbf{K}$ shine matrix na ciki na kyamara. Don $n \geq 3$, ana iya warware wannan ta amfani da algorithms kamar EPnP ko hanyoyin maimaitawa. Matsayin robot shine $\mathbf{t}$.

5. Sakamakon Gwaji & Bayanin Jadawali

Takardar ta yi iƙirarin cewa nunin gwaji ya tabbatar da daidaito mai girma da aikin gaskiya-lokaci. Yayin da ba a bayyana takamaiman sakamakon lambobi ba a cikin ɓangaren da aka ba da shi, za mu iya fahimtar yanayin sakamakon bisa ga aikin da aka ambata a baya da bayanin tsarin.

Ma'auni na Ayyukan da aka Ƙaddara:

• Daidaiton Gano Matsayi: Ambaton [2,3], wanda ya sami daidaito kusan 2.5 cm don gano matsayin robot ta amfani da LED guda ɗaya tare da SLAM, wannan tsarin haɗin kai yana da alama yana niyya ga daidaito na matakin santimita. Daidaito aiki ne na yawan LED, ƙudurin kyamara, da daidaitawa.
• Ƙimar Gano ID/Daidaito: Ma'auni mai mahimmanci don amincin tsarin. Mayar da hankali na takardar akan wannan yana nuna gwaje-gwaje sun auna ƙimar kuskuren bit (BER) ko ƙimar fassarar nasara a ƙarƙashin yanayi daban-daban (nisa, kusurwa, hasken muhalli).
• Jinkirin Gaskiya-lokaci: Jinkirin ƙarshe-zuwa-ƙarshe daga ɗaukar hoto zuwa nuna matsayi akan wayar hannu. Wannan ya haɗa da sarrafa hoto, fassarawa, binciken ma'ajin, da ƙididdigar matsayi. Don ingantacciyar haɗin gwiwa, wannan yana buƙatar kasancewa ƙasa da 100ms.

Bayanin Jadawali (Hoto na 1):
Hoto na 1 yana nuna yanayin gwaji gabaɗaya. Yawanci zai haɗa da:

1. Zane ko hoton yankin gwaji tare da masu watsa LED huɗu da aka sanya a sanannun ma'auni akan rufi ko bangon.
2. Dandalin robot (misali, robot mai tuƙi daban-daban ko mai kewayawa duk faɗin) mai ɗauke da kyamara mai fuskantar sama.
3. Mai amfani yana riƙe da wayar hannu, tare da kyamararsa kuma an nuna shi zuwa LED.
4. Ƙaramin shiga ko bangare na daban yana nuna fuskar nuni na wayar hannu, yana ganin taswira tare da gumaka masu wakiltar matsayin gaskiya-lokaci na robot da wayar hannu kanta.

6. Tsarin Bincike: Nazarin Lamari Ba tare da Lambar Ba

Yanayi: Zaɓin Oda a Cikin ɗakin Ajiya tare da Ƙungiyoyin Mutum da Robot.
Manufa: Robot yana jigilar keke zuwa tashar zaɓe inda ma'aikacin ɗan adam ke haɗa kayayyaki. Dukansu suna buƙatar bayanan matsayi daidai, da aka raba don haɗuwa mai inganci da kaucewa cikas.

Aikace-aikacen Tsarin Aiki:

1. Saitin Kayayyakin More Rayuwa: Rufin ɗakin ajiya an sanya shi da grid ɗin fitilun LED masu ikon VLP, kowanne an tsara shi da UID ɗinsa da daidaitattun ma'auni na ɗakin ajiya (misali, Layi 3, Bay 5, Tsayi 4m).
2. Gano Matsayin Robot: Kyamarar da ke saman robot tana ci gaba da kallon LED da yawa. Tana fassara IDs ɗinsu, tana dawo da matsayinsu na 3D daga taswira na gida ko na gajimare, kuma tana amfani da PnP don ƙididdige matsayinta (x, y, theta) akan bene na ɗakin ajiya tare da daidaito kusan 5cm.
3. Gano Matsayin Ma'aikaci: Wayar hannu na ma'aikaci (a cikin hular da aka ɗora a ƙirji don daidaitaccen alkibla) tana aiwatar da tsarin VLP iri ɗaya. An ƙididdige matsayinsa, amma kuma ana raba shi ta Wi-Fi zuwa tsarin tsakiya da robot.
4. Ma'anar Haɗin Kai:
   • Manajan ayyuka na tsakiya yana ba robot manufa: matsayin ma'aikaci na yanzu.
   • Robot yana tsara hanya, yana amfani da matsayinsa da matsayin ma'aikaci da aka sabunta cikin sauri.
   • A kan allon wayar hannu na ma'aikaci, wani murfin AR yana nuna matsayin rayuwa na robot da kiyasin lokacin isowa.
   • Idan ma'aikaci ya motsa, manufar robot tana sabuntawa cikin sauri, yana ba da damar sake tsarawa cikin sauri.
5. Sakamako: Rage lokacin bincike, kawar da haɗin gwiwar magana, ingantattun hanyoyi, da haɓaka aminci ta hanyar sanin juna.

7. Fahimtar Jigo & Ra'ayi na Mai Bincike

Fahimtar Jigo: Wannan takarda ba game da ƙirƙirar sabon algorithm na gano matsayi ba ce; wasa ne na haɗa tsarin mai amfani. Mahimmin darajar shine a haɗa manyan abubuwan guda biyu masu girma—kyamarorin wayoyin hannu da ake samu ko'ina da tsarin aiki na robot (ROS ecosystem)—tare da kayayyakin more rayuwa na LED don warware matsalar "haɗin kai na mita na ƙarshe" a cikin sarrafa kansa. Yana sake amfani da tashar sadarwa (haske) don amfani biyu azaman fitilar gano matsayi mai inganci, ra'ayi mai kama da ƙa'idodin haɗa na'urori da ake gani a cikin tsarin SLAM mai ci gaba amma tare da ƙarancin farashi mai yuwuwa da sarrafa kayayyakin more rayuwa mafi girma.

Kwararar Ma'ana: Hujja tana da ma'ana: GPS ya gaza a cikin gida → VLP yana ba da madadin mai yiwuwa, mai daidaito → aikin da ya gabata yana nuna nasara akan dandamali ɗaya → saboda haka, haɗa waɗannan cikin tsarin haɗin kai yana buɗe sabbin aikace-aikacen haɗin gwiwa. Kwararar daga fasahar sassa (OOK, rufe tagogi a jere) zuwa tsarin ƙarami (VLP akan waya) zuwa tsarin haɗe (tsarin raba matsayi) yana bayyana kuma yana da ma'ana.

Ƙarfi & Kurakurai:
Ƙarfi: 1) Amfani Biyu Mai Kyau: Yin amfani da haske da na'urori masu gani da ake da su yana rage farashin kayan aiki. 2) Daidaito Mai Girma Mai Yiwuwa: Hanyoyin da suka dogara da gani na iya fi tsarin da suka dogara da RF (Wi-Fi/Bluetooth) a cikin yanayi da aka sarrafa. 3) Sirri & Tsaro: Na asali na gida kuma mai gani, ba kamar bin diddigin RF ba.
Kurakurai Masu Muhimmanci: 1) Kurmin Gani Kai Tsaye (LoS): Wannan shine ƙafar Achilles. Duk wani toshewa—ɗaga hannu, pallet, jikin robot kansa—yana karya gano matsayi. Da'awar magance "yanayi daban-daban na haske" [5-7] mai yiwuwa tana magance hayaniyar hasken muhalli, ba NLoS ba. Wannan yana iyakance ƙarfi sosai a cikin ɗakunan ajiya masu cike da cikas da motsi. 2) Dogaro da Kayayyakin More Rayuwa: Yana buƙatar grid ɗin LED mai yawa, daidaitacce, da daidaitacce. Sake gyara wuraren da ake da su ba abu ne mai sauƙi ba. 3) Tambayoyin Haɓakawa: Ta yaya tsarin ke ɗaukar dozin robobi da ma'aikata? Rashin tsangwama da matsalolin binciken ma'ajin ba a magance su ba.

Fahimta Mai Aiki:

1. Haɗa ko Mutu: Don yiwuwar duniyar gaske, wannan tsarin VLP dole ne ya zama sashi a cikin tarin gano matsayi na gauraye. Ya kamata a haɗa shi da odometry na ƙafa, IMUs, kuma watakila ultra-wideband (UWB) don juriya na ɗan lokaci na NLoS, kama da yadda SLAM na Google's Cartographer ke haɗa bayanan lidar da IMU. Ya kamata a ƙera tsarin tare da haɗa na'urar gani a matsayin ɗan ƙasa na farko.
2. Mayar da Hankali kan Yarjejeniyar Musafaha: Sabon abu na takardar shine gano matsayi na "haɗin kai". Mafi mahimmancin R&D ya kamata ya kasance akan ka'idar sadarwa tsakanin wakilai—ba kawai raba ma'auni ba, amma raba tazara na amincewa, niyya, da haɗin gwiwar warware shubuha lokacin da wakili ya rasa LoS.
3. Gwada da Matsayin Fasaha na Zamani: Dole ne marubutan su yi kwatankwacin tsarin su na daidaito, jinkiri, da farashi da tsarin da suka dogara da UWB (kamar Pozyx ko tsarin AirTag na Apple) da tsarin alamar gaskiya da suka dogara da kyamara (kamar AprilTags). Bukatar darajar tana buƙatar ƙayyadaddun ma'ana.

8. Hasashen Aikace-aikace & Hanyoyin Gaba

Aikace-aikace na Kusa (shekaru 3-5):

• Ƙwarewar Ajiya & Kayan Aiki: Kamar yadda aka zayyana a cikin nazarin lamari, don daidaitaccen doki, zaɓin haɗin gwiwa, da sarrafa kaya inda robobi da mutane suke raba sarari.
• Ƙwayoyin Masana'antu na Ci Gaba: Jagorantar robobi masu haɗin gwiwa (cobots) don mika sassa ga masu fasaha a takamaiman wurare akan layin taro.
• Kasuwanci Mai Mu'amala & Gidajen Tarihi: Samar da bayanai masu sanin yanayi akan wayoyin hannu dangane da takamaiman matsayi a ƙarƙashin takamaiman hasken nunin, da kuma jagorantar robobin sabis don taimaka wa baƙi.
• Wuraren Rayuwa masu Taimako: Bin diddigin matsayin mazauna (tare da izini) da jagorantar robobin taimako zuwa gare su, yayin tabbatar da sirri ta hanyar sarrafa gida.

Hanyoyin Bincike & Ci Gaba na Gaba:

1. NLoS da Ƙarfi: Bincike kan yin amfani da tsarin hasken da aka nuna ko haɗa VLP tare da wasu nau'ikan na'urori masu gani (sauti, zafi) don ƙididdige matsayi yayin ɗan gajeren toshewar LoS.
2. Daidaituwa & Haɗin Kai: Haɓaka ƙa'idodi na buɗe ido don tsare-tsaren daidaitawa na LED na VLP da tsarin bayanai, kama da ma'aunin IEEE 802.15.7r1 don VLC, don ba da damar yanayin masu siyar da yawa.
3. Sarrafa da AI Mai Ƙarfafawa: Yin amfani da koyo mai zurfi don fassarar ID mai ƙarfi a ƙarƙashin matsanancin bambance-bambancen haske, blur na motsi, ko ɓarna na ɓangare, matsawa bayan bututun gani na gargajiya.
4. Haɗawa tare da Tagwayen Dijital: Bayanan matsayi na gaskiya-lokaci na duk wakilai ya zama cikakkiyar ciyarwa ga tagwayen dijital na rayuwa na wani wuri, yana ba da damar kwaikwayo, ingantawa, da nazarin hasashe.
5. Ka'idoji masu Ƙarfin Makamashi: Ƙirƙirar ka'idoji don wayoyin hannu su aiwatar da VLP tare da rage ƙarfin baturi, watakila ta amfani da masu sarrafa haɗin gwiwa masu ƙarancin wutar lantarki ko dubawa na lokaci-lokaci.

9. Nassoshi

1. [Marubuci(a)]. (Shekara). Take na hanyar gano matsayi na robobi dangane da ROS. Sunan Taro/Jarida. (An ambata a cikin PDF a matsayin [1])
2. [Marubuci(a)]. (Shekara). Take na hanyar gano matsayi na robot dangane da LED guda ɗaya. Sunan Taro/Jarida. (An ambata a cikin PDF a matsayin [2])
3. [Marubuci(a)]. (Shekara). Take na takardar da ta haɗa gano matsayi na LED guda ɗaya tare da SLAM. Sunan Taro/Jarida. (An ambata a cikin PDF a matsayin [3])
4. [Marubuci(a)]. (Shekara). Take na aikin da ya nuna yiwuwar gano matsayin robot na haɗin gwiwa. Sunan Taro/Jarida. (An ambata a cikin PDF a matsayin [4])
5. Zhou, B., et al. (Shekara). Tsare-tsare na VLP Masu Daidaito Mai Girma don Wayoyin Hannu na Zamani. IEEE Transactions akan Lissafin Wayar Hannu. (Misalin wallafe-wallafen tsarin VLP)
6. Ma'aunin IEEE don Gida da cibiyoyin birane–Sashi na 15.7: Sadarwar Wireless ta Gani ta Gajeren Zango. (2018). IEEE Std 802.15.7-2018. (Ma'auni mai iko don VLC)
7. Grisetti, G., Stachniss, C., & Burgard, W. (2007). Ingantattun Dabarun Taswirar Grid Tare da Masu Tace Barbashi na Rao-Blackwellized. IEEE Transactions akan Robobi. (Nassin tushe na SLAM mai dacewa don mahallin gano matsayi na robot)
8. Apple Inc. (2021). Nemo Daidaito don AirTag. [Gidan yanar gizo]. (Misalin tsarin gano matsayi na kasuwanci na UWB a matsayin ma'auni mai gasa)
9. Olson, E. (2011). AprilTag: Tsarin alamar gani mai ƙarfi da sassauƙa. Proceedings of the IEEE International Conference on Robotics and Automation (ICRA). (Tsarin alamar tushen madadin da ake amfani da shi sosai)