drayRobotics

Robotics for delivery-first grocery

Grocery.Run by robots.

DRAY connects shelf intelligence, navigation and robotic handling into a platform for delivery-first grocery.

Discuss a pilot location Watch a recorded task

Demonstrated in simulation.
Seeking a first pilot partner in Germany.

A simulated DRAY robot handling grocery products beside a stocked shelf
DRAY in simulation

Robotic picking. Intelligent shelves.

Orders in.
Groceries out.

Follow the workflow
ReplenishRelocatePick
Shelf replenishment · simulation

DRAY in action. Recorded demonstrations from our simulated store.

The opportunity

Build the capability once.
Put it to work across stores.

A retail network is more than one customer. It is many locations doing the same essential work: keeping shelves ready and getting orders out.

01 / The platform

From knowing
to doing.

Connect what is on the shelf to what the robot does next. Picking, replenishment and shelf checks become parts of the same workflow.

See the recorded workflow
02 / The first market

Delivery-first.
Germany first.

Start with one focused grocery location and a selected assortment. Design routine work around robots, with people handling support and exceptions.

Explore the pilot
03 / The scale thesis

One location.
A path to a network.

Establish the operating model at the first site. Use the evidence to define which workflows and deployment practices can repeat across a retail network.

Explore the investment thesis

The proof · Recorded in simulation

Read the shelf.
Act on what matters.

Follow one recorded replenishment task from a shelf image to a verified result. The realogram connects product identities and positions to the robot’s next action.

See → Decide → Act → Verify
From shelf data to completed work
01 / See

Every product.
A place
on the shelf.

DRAY reads product identities, positions and visible facings from the shelf image. The realogram makes that information usable for shelf audits, task selection and navigation.

Shelf recognition · recorded simulation
Shelf camera / recognitionSimulation
Shelf camera image before restockingThe same shelf image with recognized products and positions outlined
02 / Decide

One missing
facing.
One clear task.

Compare what is on the shelf with what belongs there. Here, DRAY finds two facings where three are expected and selects the position to replenish.

Shelf insight becomes a restocking task
Selected work orderSimulation
A crosshair marks the missing product position beside two coffee products
Replenish shelf
Product
SKU 223
Observed → expected
2 → 3 facings
Action
Add one facing

A precise target connects shelf understanding to the robot’s next action.

03 / Act

Intelligence,
with a pair
of hands.

The robot takes the selected product from its tray, reaches the target and places it on the shelf. Shelf intelligence becomes a physical action.

Robotic replenishment · simulation
Restocking executionSimulation

The selected product moves from supply tray to shelf.

04 / Verify

The action ends.
The check
closes the loop.

A fresh shelf image checks the result. In this recorded task, the missing facing is restored and the shelf record is updated.

Before and after, connected by a task
Post-action shelf checkSimulation
Recognition output after restocking, with three facings of product 223 now present
Target shortage clearedProduct 223 · selected recorded task
Checked

See the shelf. Do the work. Check the result.

Explore the AI behind the action

Intelligence at work

AI that gets
the work done.

Seeing the shelf. Understanding the task. Moving the product. Explore how DRAY brings AI capabilities together to do useful work.

Intelligence in action
Restore a missing product facing.
Recorded simulation view of the robot performing a shelf task
Recorded sceneSimulation
Wrist camera view from the same recorded simulation task
Wrist viewSimulation
DRAY AIConnect context to purpose
The goalThe productThe place
Illustrated DRAY mobile manipulator with an articulated arm, suction tool and product trayReplenish the shelf
Recorded shelf recognition result
The shelf, understoodProducts · Positions · Visible facings
Shelf context + task understanding
Task understanding

Give the robot a clear job.

Connect the missing facing to the product and shelf position that need attention.

Shelf recognitionTask understandingRobotic handling
Why it matters

Keep shelves ready for the next order.

Recorded in simulation

Illustrated scenarios with recorded simulation examples. Explore the idea, then watch the work.

Recorded simulation

Recorded replenishment

A selected DRAY demonstration from our simulated store.

Inside DRAY

A store that
understands itself.

Step inside the simulation. Reveal an AI view of the world, then explore the intelligence behind every action.

Store intelligence
Explore our simulation
A different way to see

Same world. New understanding.

A grocery aisle in simulation, with stocked shelves on both sides and two people ahead
Illustrated AI perception of the same aisle, with blue spatial points and amber human silhouettes
01 / World

An aisle full of possibility.

Products to handle. People to move around. A place to understand before taking the next step.

Interactive illustration using our simulation scene. AI view and route are visual explanations.

Explore the world through layers of perception.DRAY / Physical intelligence

Recorded simulation

Navigation, from the demo.

Recorded path through the store with moving agents, from one simulated navigation experiment

A recorded navigation experiment from the demo. The interactive route illustrates a mission; this image shows the recorded run.

Recorded simulation · separate shelf experiment

One gap. One action. A new shelf state.

Recorded shelf recognition before restocking SKU 223

SKU 223 · 2 observed / 3 expected facings

This recorded shelf task is a separate experiment from the store model above.

One platform. More than one way to work.

Robotic stores

Built around
the order.

A dedicated grocery format for robotic fulfillment, with a lean team supporting the operation.

Explore the German pilot
Existing stores

Alongside
your team.

Apply the same capabilities to replenishment, misplaced products and repetitive picking in conventional stores.

See the capabilities
Shelf analytics

Know what
needs attention.

Read product positions and visible facings. Identify gaps, compare shelf layouts and check completed work.

Explore shelf intelligence

The robotics platform

Pick. Replenish.
Put things right.

One mobile platform connects shelf understanding with product handling. Explore the capabilities in our simulated store.

Replenishment · Simulation00:41

The robot approaches the shelf, takes a product from its supply tray and places it on the shelf. From a supply tray to the right shelf position.

Product relocation · Simulation00:13

The robot grasps a product, moves it along the shelf and releases it in a different location. Put misplaced products back where they belong.

Order picking · Simulation00:31

The robot reaches for a jar, draws it out of the shelf and transfers it into the tote. The core movement behind robotic order fulfillment.

Recorded DRAY demonstrations · Simulation

Our first physical pilot · Germany

One store.
A new way to operate.

We’re seeking a retail operating partner for our first physical pilot in Germany. The goal: a delivery-first grocery location where robots handle routine shelf and order work, with people supporting operations.

  1. Build around orders.

    Start with a focused assortment and a store layout designed for robotic movement, picking and dispatch.

  2. Put robots to work.

    Connect navigation, shelf intelligence and manipulation. Shape scheduled on-site maintenance, remote oversight and exception handling around the operation.

  3. Establish the model.

    Measure order fulfillment, availability, human involvement and cost. Use the results to define a repeatable store format.

Simulation → Physical pilotToday: recorded shelf recognition, picking and shelf operations in simulation. Next: integrate and validate the workflows in a physical pilot.

A pilot with a business test.

Agree success thresholds together. Measure the operation, then decide what scales.

Cost per completed order
Include equipment, support and human intervention alongside robotic execution.
Availability & completion
Track shelf availability and whether customer orders can be completed reliably.
Human involvement
Measure maintenance, exceptions and the support needed across the service window.

The first partner

Help define
the next store format.

Bring your retail experience. Build with us from the first location, and help shape how a successful model can expand across a network.

The first conversation covers the site, assortment and operating goals. Together, we define responsibilities, pilot funding and success thresholds before moving forward.

Discuss a pilot location

Why join early

  • Shape the formatHelp set the assortment, workflow and operating priorities around a real retail business.
  • Build practical knowledgeLearn the economics and day-to-day requirements of a robotic store through a shared pilot.
  • Create a path to more locationsUse the first site to establish what transfers across your network, with each expansion agreed on its own merits.

For early-stage investors

The store is the pilot.
The platform is the ambition.

DRAY is building the robotics platform connecting shelf understanding, movement and manipulation. A focused grocery operation is the first place to validate it end to end.

We develop and test connected AI capabilities in simulation, with recorded work to inspect. The next milestone is to validate the operation in a physical pilot.

Discuss the company & next milestone
  1. 01 / Demonstrated in simulation

    Connect the workflow.

    Shelf recognition, replenishment, relocation and picking, with recorded tasks to inspect.

    Explore the demonstrations
  2. 02 / Next milestone

    Prove the operation.

    Validate integrated physical workflows, support needs and operating economics with a pilot partner.

  3. 03 / Scale ambition

    Make deployment repeatable.

    Turn what works at the first site into a format for further locations, guided by operating evidence.

A few practical questions

Before we
get to work.

Have a specific store or workflow in mind? Tell us about it.

What can I see today?

The footage shows DRAY’s shelf recognition, replenishment, product relocation and picking in simulation. Our next milestone is to bring the capabilities together in a physical store pilot.

Is this only for stores run by robots?

The delivery-first store is our flagship pilot concept. The same platform also targets shelf work alongside store teams and shelf analytics that help people decide what needs attention.

What does the realogram tell us?

It structures the visible shelf: product identities, positions and facing counts. Compared with the expected layout, that supports gap detection, misplaced-product checks and verification after a task. It also provides product landmarks for navigation.

What role would people play?

The pilot concept puts people in charge of maintenance, oversight and exceptions while robots handle routine shelf and order work. The first site will establish the right balance of scheduled visits, remote support and on-site intervention.

Could the store operate on Sundays?

Extended service windows, including Sundays where permitted, are part of the concept. Opening, staffing and delivery arrangements depend on the location and operating model; these would be assessed with the pilot partner.

How do we explore a partnership?

Start with a potential location, assortment and your retail goals. We’ll discuss site fit, operating responsibilities, support, pilot funding and success measures together. The commercial structure and any wider rollout would be agreed directly.

Let’s take the next step

See a place
for DRAY?

Retailers & operators

Explore whether your location and operating goals fit the first Germany pilot.

Discuss a pilot location

Early-stage investors

Discuss the platform, the evidence so far and the next company milestone.

Start an investor conversation