Project support · for partners and funders · updated 2026-10-28

Support Arabati Wein

عربتي وين؟ means “where is my car?”. It is a free, scam-free public registry of vehicles looted or lost in the Sudan war, live at mycarsd.online. One Sudanese family built it while searching for their own car. This page sets out what it would take to run it for every family, and what each level of funding adds.

First goal: keep the registry running when nobody is at the keyboardUS$0 of US$3,000
No donation route is open yet. A fiscal sponsor or registered non-profit comes first (see “What happens next”). This page is the plan, not a collection.

ملخّص بالعربية

«عربتي وين؟» سجلٌّ عام مجاني للعربات المنهوبة والمفقودة في حرب السودان، على الرابط mycarsd.online. يجمع كشوفات العربات التي تنشرها لجان الشرطة والمتطوعون والمجموعات العامة، نصاً وصوراً، ويحوّلها إلى سجل واحد لكل عربة يمكن البحث فيه برقم الشاسي أو اللوحة بالأرقام العربية أو الإنجليزية. من لا يجد عربته يسجّل رقمها، ويُفحص الرقم كل ساعة مقابل كل منشور جديد، ويُبلَّغ صاحبه إن ظهر. كل صفحة تُرشد إلى الطرق الرسمية المجانية (البلاغ الإلكتروني في sudanpolice.net وإدارة المرور) وتحذّر من النصب: لا يوجد رقم واتساب رسمي، ولا رسوم للاسترداد، ولا ترسل رقم شاسيك في الخاص لأحد.

المشروع بناه أفراد أسرة سودانية واحدة أثناء بحثهم عن عربتهم، ويعمل حتى اليوم بلا تمويل على حاسوب منزلي واحد. هذه الصفحة تبيّن ما يلزم لتشغيله لكل الأسر: أولاً خادم دائم وتنبيهات آلية (٣٠٠٠ دولار)، ثم معالجة يومية وروبوتات واتساب وتيليجرام (١٠ آلاف)، ثم منسّق مجتمعي ودليل الأحواش الرسمية (٢٥ ألفاً)، ثم مؤسسة مسجَّلة تتعاون مع شرطة السودان وإدارة المرور (٦٠ ألفاً). المبادئ لا تتغير بالمال: مجاني دائماً، التنبيه برقم مطابق لا بالشكل، لا تواصل مع البائعين، والأرقام تُنشر بصدق بما فيها صفر استردادات حتى الآن.

Why this exists

When Khartoum emptied in 2023, cars were taken by the hundreds of thousands. They are now found in police yards, photographed by field volunteers, listed by recovery committees and offered for sale in dealer groups, in lists that are posted as photos of handwritten pages, scattered across dozens of Facebook groups and Telegram channels. A family that still holds its chassis number has nowhere to type it.

Into that gap came extortion. Fake “recovery committees” ask for the chassis number in private, demand fees, hijack WhatsApp accounts through the six-digit code, then vanish. The registry answers both problems: one place to search, and a clear route to the official free channels, with a scam warning beside every contact number it shows.

The demand is documented by the state. Sudan Police’s electronic reporting platform logged 107,656 reports in June 2026 alone, 78,464 of them missing vehicles, according to the police spokesman’s public statement. Khartoum State committees hold more than 8,000 collected vehicles at 38 sites. Owners are told to come and search by chassis number.

Who runs it

A Sudanese family, as volunteers, with a personal AI subscription and under US$200 of direct costs since August 2026. There is no legal entity yet. The code is published under the Apache 2.0 licence so the project can be audited, contributed to, and run by others; the policies it follows are public (privacy, safeguarding) and do not change with money.

What exists today

78,379distinct chassis numbers searchable
93,584identified vehicles (distinct chassis, or plate where no chassis)
159,310posts collected since 22 Aug 2026 from 46 groups, 2 pages and 5 channels
48,257photos transcribed by OCR so a stamped VIN becomes text
39standing watches registered by families who found nothing
2leads relayed to owners by hand
0confirmed recoveries so far. Published honestly every month
110,474posts collected but not yet processed (the backlog)

Measured from the project database; figures refresh from stats.json on each visit and in full at /stats. We never quote raw sighting rows as vehicles, because the same list is reposted to many groups.

Built and working

  • Arabic-first search with English toggle, by chassis, plate, make, model or year, in Arabic or Latin digits.
  • Collectors for Telegram channels and public Facebook recovery groups; a Telegram bot families forward posts to.
  • A parser for the long handwritten كشف lists, plus OCR of every photo, with a grounding gate that deletes any identifier not present in the source. Over a hundred invented numbers were caught this way; none were self-reported by the models.
  • Alerts only on an exact identifier, never on resemblance. A white 2015 Corolla looks like every other one.
  • Standing watches: a family that finds nothing leaves a number and is checked against every new list, hourly.
  • Official free channels and the scam warning on every page, including the “no official WhatsApp number exists” rule.
  • Visitors from Sudan, Egypt, the Gulf and the diaspora, almost all on phones. AI assistants have begun recommending the site.

Not yet, and why it matters

  • It runs on one home computer. When the computer is off or short of memory, nothing is collected and no watch is checked. It was off for three days in October.
  • Lead relay is still by hand. A matched watch alerts the operator and prepares the message; a person still sends it.
  • Intake outruns processing. Three to five thousand posts arrive a day; the free processing tier reads about eight hundred. The backlog grows daily.
  • Lists are scraped, not sent. The sustainable model is recovery groups, field platforms and police committees sending their lists directly.
  • No messaging-app access. Sudan visits the website rarely; Sudan lives on WhatsApp and Telegram.
  • No legal entity. Nobody can receive a grant, sign a data-sharing agreement, or take over if the founders cannot continue.

How much it costs

Each rung includes everything below it. Amounts are twelve-month figures built from unit costs the project has measured: about US$0.006 to read one post, US$1.50 to transcribe a thousand photos, about US$40 a month for an always-on server. A machine-readable copy is at funding.json.

  1. Running now, unfunded

    US$0

    Free tiers, one volunteer, a personal AI subscription, and under US$200 in direct costs since August.

    • Everything listed under “Built and working”, at the pace one person and a home computer allow.
  2. Never offline

    US$3,000

    The registry stops depending on one laptop, and a family is told when its number appears.

    • Move collectors, the hourly watch check and daily processing to a cloud server with monitoring and backups. Two years of hosting and domain included.
    • Automatic delivery of the lead message to a watch registrant, by Telegram or email, when a new list carries their chassis or plate.
    • Clear the backlog once, so the site reflects everything collected.
    • Security review of the published code; dependency and secrets audit.
    • Published privacy and safeguarding policies, kept current.
  3. Same-day registry

    US$10,000

    Every list is searchable the day it is posted, and people can search from the apps they already use.

    • Paid processing that keeps pace with intake, so the backlog stays at zero. About US$540 a month at current volume.
    • OCR for every new photo. Stamped compliance plates in photos are the richest source of full VINs and are invisible to text search.
    • WhatsApp and Telegram search bots using the same matching rules as the website.
    • A low-data version of the site for Sudanese mobile networks. Today the data payload is close to 10 MB.
    • Monthly statistics on /stats: distinct chassis, watches, leads relayed, recoveries confirmed.
  4. People on the ground

    US$25,000

    Someone whose job is to verify leads, relay them safely and answer families in their own dialect.

    • A part-time Sudanese Arabic community lead for twelve months.
    • A verified directory of official impound yards and collection sites, with a plain guide to the claim procedure: identity document, proof of ownership, the court-issued sworn affidavit. Three government facilities have already been identified from photographed signage; all three are also used as scam bait.
    • Direct intake agreements with recovery-group admins and field platforms, so lists arrive from the source instead of being scraped.
    • Owner photo matching for cars whose chassis plates have been swapped: distinguishing marks, dents, stickers, a phrase on the rear window. A WhatsApp group in Khartoum recovered twelve cars this way.
    • A legal-aid referral path with the Ministry of Justice Legal Aid Administration and its partners.
  5. An institution, not a laptop

    US$60,000

    A registered non-profit that can sign agreements with the authorities and survive its founders.

    • Registration as a non-profit with a board, annual accounts and an independent data-protection review.
    • One engineer and one community lead for a year.
    • A data-sharing agreement with the Sudan Police reporting platform and the Traffic Department, so a registry match can feed the legal registration ban and INTERPOL listing that only an official report triggers.
    • An open API and data exchange with the other registries, including the police committees’ own lists.
    • Registration as a Digital Public Good and an independent impact evaluation.
  6. Beyond Sudan

    US$150,000

    The same problem exists in every war. The pipeline becomes a template other communities can run.

    • Three years of sustained operation for Sudan.
    • A documented, deployable version for partner organisations in other conflict-affected countries, with the fabrication gates and scam safeguards built in.
    • Multilingual interface and list parsers.

Twelve-month budget

LineBasisUS$ / year
Always-on server, monitoring, backupsabout US$40 a month480
Processing to keep pace with intake90,000 posts a month at a measured US$0.006 each6,500
Photo transcriptionabout 60,000 photos a month at US$1.50 per thousand1,080
Messaging bots and notificationsWhatsApp Business API, SMS gateway1,200
Hosting, domain, security review300
Community lead, part timetwelve months9,600 – 14,400
Engineer, half timetwelve months24,000 – 36,000
Fiscal sponsor or registration, accountstypically 10% of funds—
Minimum programme, no salaries≈ 9,500
Staffed programme≈ 50,000 – 60,000

What is needed now

  1. A fiscal sponsor so grants and donations can be received before a non-profit exists. Open Source Collective or an equivalent host.
  2. A Sudanese civil-society co-applicant for humanitarian funds that require a lead registered in the project country.
  3. Data partners: recovery-group admins and field platforms willing to send lists directly.
  4. In-kind cloud credits and a first grant in the US$3,000–10,000 range to take the registry off a home computer.
  5. Introductions to the Traffic Department and the Sudan Police reporting platform for a data-sharing conversation.

What funding will not do

It will not recover a specific car, and it will never charge a family or contact a seller. The registry tells a family where a number has appeared and sends them to the police, the Traffic Department and the official yards. Approaching a holder directly is how families get extorted, and the project will not do it or help anyone do it.

We also do not promise matches. A chassis search can fail when a looted car has been given a wrecked car’s stamped parts, which is the standard laundering method. That is why photo matching and official data sharing are on the ladder.

Principles that do not change with money

How to help now

Recovery groups and field platforms

Send us your lists directly and we will credit you on every record and give your members a search link. This replaces scraping.

Funders and NGOs

We are seeking a fiscal sponsor and a Sudanese civil-society partner to co-apply. A one-page concept note and the budget above are ready on request.

Engineers and Arabic speakers

The code is open source under Apache 2.0. Arabic-digit search, list parsing and messaging bots are the first open issues.

Lawyers and officials

Help us write the claim guide and open a data-sharing conversation with the Traffic Department and the police reporting platform.

Questions funders and families ask

Is Arabati Wein free to use?

Yes, always. There is no fee to search, to register a standing watch, or to receive a lead. Official vehicle recovery in Sudan is also free; anyone asking for money to release a car is not the police.

Will the registry recover my car?

No. It tells you where your chassis or plate number has appeared in public posts and sends you to the official channels: the free report at sudanpolice.net/reg/car_stealing and the Traffic Department in person. It never contacts sellers or holders.

Do you contact the person who posted my car?

Never, and we ask families not to either. Approaching a holder directly is how families get extorted and how a car disappears again. The identifier and an official report are what settle it.

What do you store about me?

If you register a watch: the number you typed, your contact and your note, used only to tell you when the number appears. Never published. The site records search queries in analytics; see the privacy page. Photos and text read from photographed documents are never published.

How are the numbers in the registry verified?

Every identifier must be present in the source post’s text or in the OCR of its photos, or it is deleted. Alerts need an exact chassis or plate match; resemblance is never enough. Each record links to the original post so you can check it yourself.

How can an organisation fund or partner with the project?

We are seeking a fiscal sponsor, a Sudanese civil-society co-applicant, data partnerships with recovery groups and field platforms, and grants between US$3,000 and US$60,000. The funding ladder above states what each level adds. Contact details are below.

What happens next

Ninety days to be ready for funding
  • Weeks 1–2. Open the code, publish privacy and safeguarding policies, publish the statistics page, build watch notifications. Done, October 2026.
  • Weeks 3–4. Secure a fiscal sponsor so donations and grants can be received; begin non-profit registration; finish the twelve-month budget and concept note.
  • Weeks 5–8. Apply to civic-technology and humanitarian funds; submit for Digital Public Good registration; request in-kind cloud credits; open partnership conversations with group admins and field platforms.
  • Weeks 9–12. Spend the first funds on the cloud migration and the messaging bots, and publish the first monthly impact report.

Contact

Project contact: firasfmohammed@outlook.com

For families: the official free report is at sudanpolice.net/reg/car_stealing; we do not take reports on behalf of the police and we have no WhatsApp number. For organisations: say who you are and which rung of the ladder or which partnership you are writing about, and we will reply with the concept note and budget.