Miele's G8 Diamond Dishwasher Uses a Camera and AI to Read the Load Before Washing

Miele has built a camera into a dishwasher to identify what is inside before the wash starts. The German appliance maker showed the G8 Diamond with AutoVision at IFA 2026. When the door closes, the machine photographs the tub and uses AI image recognition to identify the load, whether that means wine glasses, greasy pans or plastic containers. It then adjusts the wash to suit.

Dishwasher programmes have become more automatic over the years, with sensors measuring water cloudiness and temperature. AutoVision adds something those sensors cannot provide: an understanding of what the items in the racks actually are.

What AutoVision Does

• Snapshot on closing: an integrated camera photographs the tub when the door shuts.
• Load recognition: AI models identify items including wine bottles, greasy pots, delicate glassware and plastic containers.
• Cycle selection: the dishwasher adjusts the wash programme based on what it detects.
• Detergent dosing: it controls the amount of detergent used through Miele's AutoDos system with PowerDisk.
• Drying: it can switch on ExcellentDry heated-air drying when the load needs it.

The practical benefit is clear. Plastic tubs that come out wet, delicate glasses that need gentler treatment and roasting tins that need more heat often call for different settings. Most households pick one programme and put up with the compromise. AutoVision is designed to make that choice automatically for each load.

A dishwasher that recognises its load is not just a gimmick. It solves a problem most people have simply learned to live with.

Privacy by Default

Putting a camera inside a kitchen appliance raises obvious questions, and Miele has addressed them in one important way: AutoVision is switched off out of the box. Owners have to enable it manually. That is a sensible choice when connected home cameras are under growing scrutiny, even if the camera in this case points at crockery rather than people.

Buyers will still want clear answers about where images are processed, whether any leave the machine and how long they are kept. Miele will need to explain those details clearly before the product goes on sale.

The Smart Home Gap

The G8 Diamond works with the Miele app. However, neither Miele's announcement nor the coverage from the show mentions Matter, Zigbee or any other route into mainstream smart home platforms. That puts it behind the direction the rest of the industry is taking at IFA, where Matter support has been widespread.

• What works: app control and Miele's own ecosystem.
• What is missing: a confirmed route into Apple Home, Google Home, Alexa, SmartThings or Home Assistant.
• Why it matters: households with a mixed smart home cannot yet include the dishwasher in wider automations, such as running it when electricity prices are low.

Availability

Miele has confirmed a launch in Germany, Austria and Switzerland in April 2027. No price has been announced for any market, and there is no confirmed date for the UK or US. The G8 name also indicates a premium product at the top of Miele's dishwasher range, so buyers should not expect camera recognition in entry-level models soon.

The Bigger Trend

Cameras and image recognition are moving into large appliances. Refrigerators already use them to track food, and some ovens recognise dishes. Dishwashers are a logical next step because the load changes every time and the right settings depend on what is inside. Miele's reputation for durable, long-lasting appliances gives the idea some credibility with buyers who might otherwise dismiss it as a gimmick.

The Bottom Line

AutoVision is a useful application of AI in a common household appliance. It chooses the programme, dose and drying setting based on what is actually in the machine, and it is off by default. The limitations are its narrow smart home support, no confirmed price and a first launch in April 2027 that is restricted to German-speaking markets. Other markets will have to wait.

Image: Matti Blume via Wikimedia Commons, CC BY-SA 4.0