AI for Life
Being in the field of Computer Vision, I am a keen observer of how things are done everywhere I go, and I see numerous situations where we waste time that might be automated using AI and existing cameras. saving a vital human hour. To share my thoughts and ideas, I am starting this series AIforLife automation and insights making our lives better every day.
The Problem at a fuel station
It was a few days ago when I visited the Ghuma gas station to get some gas. While I waited in the long line at the CNG station, like always. Because it’s the only gas station in a 10-mile radius that has a CNG installation, many individuals choose to wait 30-40 minutes for their turn or just drive on by choosing to move to another fuel pump to get their gas-filled.
In addition, the petrol & diesel station appeared less congested than CNG sites made me wonder from a customer stand point think why not just open another CNG counter aren’t they losing business but
From Business owner :- It is challenging to take decisions
- Either Fuel Pumps are idle or they are overcrowded. But there is no way to know the ground truth.
- There isn’t one formula to solve this as each Fuel Pump is in a unique location serving the local economy thriving on a different mix of (Petrol-Diesel-CNG ) vehicles
- As a business leader managing 15k+ Fuel Pump Network – How to solve this problem?
How AI can aid the Business owner with the right insight and thus also improve customer experience
While fuel stations are extensively regulated by many regulatory norms and rules, they operate on extremely tight margins and are under pressure to increase profits and sustainability.
Fueling stations now offer a variety of fuels, including diesel, gasoline, compressed natural gas (CNG), and EV charging stations. By measuring, the insight will also reveal which one is in higher demand.
All fuel stations already have security cameras, which can be improved by making them smarter with AI and DL. They can be utilized to improve customer experience, employee behavior, outliers, and vehicle traffic analysis, which will finally lead to increased sales.
KPIs that can be measured
- Traffic and operational efficiency
- In count of No. of Vehicles and Vehicle type
- Vehicle wait time
- Vehicle Fueling time
- Fuel counter Idle time
- Queue Length
- Staff behavior monitoring
- Proper queue formation to serve the customer faster
- Average time per staff in serving customers – from filling Gas to payment
- Staff inattentiveness
- Identifying Outliers
- Reason why certain vehicles take more time than average time.
- Incident Alerting
- Sparks/smoke/flames
- Smoking cigarette
- Tank filling as per protocol
- Accumulation of vehicles
Measurement of these KPIs can easily be automated using IP Camera & AI Software using deep Learning to give insight into the ground reality.
Let’s understand the impact of this with a scenario
A Fuel Station A with 1 CNG & 10 Petrol + Diesel Fuel Counters, There are often long queues for CNG but they have no space to add another counter, customers are often frustrated due to long wait times.
By measuring the additional KPI highlighted with AI and DL they can experience an increase in revenue up to 39%.
| KPI’s Measured | CNG | Petrol / Diesel |
|---|---|---|
| Number Of Counters | 1 | 10 |
| Average Wait Time | 10 Minutes | 5 Minutes |
| Vehicle Queue Length | 10 Vehicles | 2 Vehicles |
| Fuel Service Time | 5 Minutes | 5 Minutes |
| Average Order Value | ₹600 | ₹1000 |
| Total Fueling Capacity Per Day Per Counter (10 working hours per day) |
120 Vehicles | 120 Vehicles |
| Counter Idle Time (Time when no vehicle is served) |
2 hours | 5 hours |
| Counter Capacity Utilization | 80% | 50% |
| Max Avg. Revenue Per Day Per Counter Capacity | ₹72,000 | ₹1,20,000 |
| Actual Vehicles Served Per Counter | 120 | 84 |
| Actual Revenue Per Day Per Counter | ₹72,000 | ₹84,000 |
| Revenue (including all counters) |
– | – |
| Total Revenue Per Day | ₹9,12,000 | |
| Additional Revenue – If we convert unutilized 50% capacity to CNG counters | ||
| Revise No. of Counters | 6 | 5 |
| Additional Revenue Per Day | ₹3,60,000 | |
| Total Revenue Per Day | ₹12,72,000 | |
| % Increase In Revenue | 39% |
When we look at this holistically across 20k Fuel Stations, the Increase % of Revenue may vary based on local economics but, it is clearly evident that there is a definite business benefit resulting in improvement in the life of customers and Business
With all this information available across all fuel stations, oil & gas companies can make informed data-driven decisions – Like opening a new pump or introducing a new CNG spot. #AIforlife has tremendous potential to improve consumer experience by creating an efficient and safe ecosystem while increasing the business for the owner.
Such projects are now in action at two gas stations of Shell in Thailand and Singapore. AI for life has immense value in automating such mundane and repetitive instances that can derive intelligence and make it a little more convenient for us.
I would love to hear your thoughts on this. Share your comments and ideas for such instances that can be automated for the better good of life and business.