AI Analytics for the physical world.

Go beyond counting. Understand how people move, gather, and spend time in your spaces, using your existing cameras and infrastructure.

A shopping center concourse seen from above, with a restaurant's seating on the left and shoppers walking across the floor. Every visitor is marked, recent movement is traced, a family is recognized as one group, and time at a pop-up kiosk and in the seating area is measured. Faces are blurred.

Concourse · todayvs last week

Visitors2,406↑ 6%
Came in groups38%↑ 2 pts
Average visit24m 10s↑ 3%
Peak hour14:0058 in now

Trusted by global operators across industries

Lacoste
Burberry
Converse
GANT
AT&T
DHL Group
Turkcell Kuzey Kıbrıs
The Celtic Collection

Know your traffic, hour by hour.

Understand visitor volumes, group sizes, and peak hours throughout the day and across your locations.

A phone store seen from inside, with the glass entrance at the back where every visitor is counted as they come in
Entrance· in and out counted

Today's Metricsvs last week

Peak occupancy346 of 350at 15:00
Average visit duration 24 min↑ 3%
Most common age25–3445%
Under 25 27%25–34 45%35–44 12%45+ 16%
GenderWomen 62%Men 38%

People Counting

Came in2,406↑ 6% vs last Tuesday
Left2,192Average visit 24 min
Still inside214Of 350 capacity
In and out by hourToday · Entrance A
08:00 · 71 came in09:00 · 76 came in09:00 · 69 left10:00 · 93 came in10:00 · 74 left11:00 · 136 came in11:00 · 90 left12:00 · 208 came in12:00 · 132 left13:00 · 286 came in13:00 · 202 left14:00 · 320 came in14:00 · 277 left15:00 · 289 came in15:00 · 310 left16:00 · 229 came in16:00 · 280 left17:00 · 202 came in17:00 · 222 left18:00 · 191 came in18:00 · 196 left19:00 · 142 came in19:00 · 185 left20:00 · 91 came in20:00 · 138 left21:00 · 72 came in21:00 · 17 left08:0011:0014:0017:0020:00

Visitor Trend

Busiest daySaturdays18.2% of weekly visitors
Busiest weekend hour14:00–15:0033.7% of visits on weekends
Busiest weekday hour17:00–18:00After-work peak, Mon–Fri
Average visitors per hourLast 4 weeks

Spot the trends walking through your door.

Get to know your visitors like never before. Discover emerging trends with a new level of detail.

Advanced Observed Attributes

White sneakers
White sneakers31%↑ 4 pts
Denim jackets
Denim jackets14%↑ 2 pts

Discover detailed visitor attributes, uncover emerging trends, and understand how your audience evolves over time.

Knitwear
Knitwear12%↑ 3 pts
How people arrived Number of people
In groups38%↑ 2 pts
Typical group2 people
What they're wearing and carrying
Backpack18%↑ 2 pt
Hat or cap12%0 pts
Tote bag9%↓ 1 pt

Customer Persona

Gender × age range206 customers · today, 14:00 to 15:00
Gender × age range0–1213–1718–2425–3435–4445–5455–64Total
Women11225922148127
Men003233311079
Total11549225258206
Women · 25–34: 59 (28.6%)Men · 25–34: 33 (16.0%)Men · 18–24: 32 (15.5%)Women · 18–24: 22 (10.7%)Women · 35–44: 22 (10.7%)6 smaller segments: 38 (18.4%)206customers
  • Women · 25–345929%
  • Men · 25–343316%
  • Men · 18–243216%
  • Women · 18–242211%
  • Women · 35–442211%
  • 6 smaller segments3818%

Go beyond basic demographics. Explore detailed visitor segments, uncover relationships between attributes, and build a deeper understanding of your audience.

Discover what captures attention.

Reveal where people spend time, what draws them in, and which areas go overlooked.

Dwell Time & Heatmaps

Overhead camera view of a store with display tables, seating and a wall display
Display tables6m 40saverage dwell

Most handled demo devices · today

  1. 1Flagship phone42
  2. 2Foldable phone29
  3. 3Tablet18
Seating9m 10saverage dwell
Wall display0m 26smost walk past
Fewer visitsMore visits

Keep every line moving.

Track wait times, identify bottlenecks, and improve the flow of people across your locations.

A quick-service restaurant with self-order kiosks and a queue lane, where every wait is measured

Today's Metricsvs last week

Average wait2m 10s↓ 20s
Waited under 3 min84%↑ 6 pts
Peak-hour wait time4m 20s↑ 30s
Longest queue12 people↑ 4
Queue abandonments8↓ 2
In queue now5 peopleEstimated wait 2m 40s
Queue lane

Queue Times

Average wait2m 10s↓ 20s vs last week
Longest wait4m 20sAt 12:00
Over target2 hours11:00 to 13:00
Average wait by hourToday · Tills
Target 3 min 4m 20s at 12:00 08:0014:0020:00
EdgarRecommendation

Open a third till from 11:00 to 13:00. That is when waits ran over the 3‑minute target today, peaking at 4m 20s at 12:00. Waits come close to target again around 18:00, so keep a till on standby then.

Service Utilization

Service Hours12h
Staff Coverage9h
Active Service Time6h

Staff present 75% of the day, actively serving for 67% of that time.

Ask about your data.

Ask in plain language. Edgar explains your data and reveals opportunities to improve.

Analysis with Edgar

You

Why did conversion drop at Eastgate last week, and what should we change?

Edgar

Conversion fell from 24% to 19%, despite an 8% increase in visitors. Longer evening queues appear to be a key factor. Wait times doubled after 17:00, and 1 in 5 customers left the queue before being served.

Recommended Action: Consider opening a third till from 17:00 to 19:00. Queen Street uses additional coverage during similar evening traffic and achieves 25% conversion.

Try askingWhich displays draw the most attention?Where do visitors spend the longest?Which location converts best on weekends?
Message Edgar…
Conversion19%↓ 5 pts vs week before
Wait after 17:004m 10s↑ 2m 05s vs week before
Left the queue1 in 5After 17:00
Average wait by hourLast weekWeek before
Target 3 min 4m 10s at 18:00 12:0016:0020:00
Conversion by dayLast week
Week before 24%23%Mon22%Tue22%Wed19%Thu17%Fri15%Sat16%Sun

Designed with security and privacy you can trust.

Edge Signal combines reliability, control, and protection to help businesses deploy and scale Physical AI with confidence.

Enterprise & Security
AICPA SOC for Service Organizations mark

SOC 2

Independent assessment of security controls against the AICPA Trust Services Criteria.

GDPR

GDPR Compliant

Data processing practices and privacy controls designed to support GDPR requirements.

Frequently asked questions about AI Analytics

Can’t find what you’re looking for?

Contact us

What is AI Analytics?

AI Analytics is Edge Signal’s AI video analytics for physical spaces. It turns existing camera feeds into real-time insight about how people move, gather, and spend time in stores, branches, restaurants, and venues, so businesses can understand traffic, attention, and queues across every location.

What can Edge Signal AI Analytics measure?

Edge Signal AI Analytics measures people counts, occupancy, visit duration, group sizes, estimated age range and gender, observed attributes such as clothing and accessories, dwell time, heatmaps, queue length, wait times, queue abandonment, and service utilization, for a single location or across an entire network.

How does AI people counting work?

AI people counting uses existing cameras to count people coming in and going out, hour by hour. Unlike traditional door counters, Edge Signal also measures live occupancy against capacity, visit duration, group sizes, peak hours, and weekly visitor trends, giving a fuller picture of how a space is used.

What are dwell time analytics and heatmaps?

Dwell time analytics measure how long people spend in a specific area, such as a display table or a seating zone. Heatmaps show where activity is concentrated and which areas people walk past. Together, they reveal which displays and zones capture attention and which are overlooked.

Can AI Analytics measure queue wait times?

Yes. Edge Signal measures queue length, average and peak wait times, time over target, and how many people leave the queue before being served. It also shows staff coverage and active service time, and Edgar can recommend actions such as opening another till during peak hours.

Does AI Analytics identify individual visitors?

No. Edge Signal AI Analytics reports on visitors as groups and segments, such as estimated age range, gender, group size, and observed attributes, without identifying individuals. Edge Signal does not collect or store personally identifiable information, and faces are blurred by default.

Does Edge Signal work with existing CCTV cameras?

Yes. Edge Signal works with compatible existing CCTV and security camera systems, using on-site Edge AI hardware to process video locally. Businesses can add AI Analytics without replacing their camera infrastructure. Camera compatibility and coverage are assessed during deployment planning.

Can I compare analytics across multiple locations?

Yes. Edge Signal brings analytics from every location into the Physical AI Platform. Teams can compare traffic, conversion, dwell time, and wait times between locations, follow historical trends, and benchmark each location against the best performer to see where improvements will have the most impact.

Can I ask questions about my analytics in plain language?

Yes. Edgar, Edge Signal’s Physical AI model, answers questions about your data in plain language, such as “Why did conversion drop at this store last week?” Edgar explains the result, connects metrics across locations, identifies likely causes, and recommends what to change.

What is the difference between AI Analytics and Physical AI Agents?

AI Analytics shows patterns over time, such as traffic, attention, and wait times. Physical AI Agents act in the moment: they watch for specific conditions, alert the right person, and track the follow-up. Used together, AI Analytics shows what to improve and AI Agents help teams act on it.

See more with the infrastructure you already have.

See what your existing cameras can reveal about visitor behavior, traffic patterns, and how your spaces are used.