Inspiration

Kwant.ai is inspired by our team's personal experience. Niran's 10+ experience as project manager and data analyst in large-scale construction projects in 3 continents and 6 countries including 2nd Avenue Subway and East Side Access combined with Sagar's experience building data analytics software using time scale data.

In the US, 14 workers die in construction sites every day which is the highest in any industry and its increasing. ( Based on American Federation of Labor and Congress of Industrial Organizations). Workers compensation claims are lowering construction companies EMR ( Experience Modifier Rate) due to a lack of data-driven safety approach. Insurers are losing $40B in claims.

Similarly, construction productivity has been declining for the last decade. With labor making 45% of the construction cost, profit margins are depended upon the workers' productivity.

Like Fitbit enables real-time activity monitoring for fitness, we build this product to show how field data collection can be automated using sensors. Using latest BLE and Lpwan Technologies we can get accurate location data in real time.

What it does

Kwant.ai is a BLE - Lpwan based sensor installed on Hardhat which collects time, location and safety data of each field workers at the construction site via gateway installed on each floor and entrances. Comparing planned vs actual progress and man-hours, accurate production unit rates are calculated enabling optimization of project schedule and proactive decisions in real time. With automated safety notifications, including who, when and where, kwant.ai adds clarity and precision to currently inefficient safety documentation, and responses processes.

How we built it

For the hackathon, we used our existing platform "onTarget" and integrated with BLE sensors on installed on Hardhats. We used gateways installed on various location in the job site to transmit the data from hardhat to AWS. In AWS IoT. AWS iot is used to transfer data from beacon to aws dynamo database via gateway using mqtt protocol. aws lambda to listen to changes in dynamo database and trigger a push to postgres database with changes.

Challenges we ran into

We ran into the challanges of multiple gateways in the job site communicating with the sensors at the same time. For the accurate location reporting, we needed to use the rssi data " received signal strength" to determine the proximity. We also worked with Lorawan technology which covers 5-12miles of the area and just using battery-powered sensors, we are able to find the proximity of the workers accurately.

Accomplishments that we're proud of

We are proud of what we have built in a short period of time which not only helps construction companies build more efficiently but as save lives.

Technically, we were able to use the combined power of BLE and Lpwan for the first time which is very important in dynamic construction jobsite.

What we learned

1.Safety is the most important metrics for construction workers and project managers spend 1-2hours a day in headcount and preparing daily reports.

  1. Advanced BLE and Lpwan ( Low Powered Wide Area Network ) technology and most importantly its hybrid technology is going to be a breakthrough.
  2. It can not only be used for headcounts and safety but also for progress tracking, payroll verification and insurance using predictive analytics/Machine learning.

What's next for Kwant.ai

We are piloting the sensors on 100 workers in 4 jobsites alread and plan to make a full launch in 1Q2019. The results and the feedback has been great.

We plan to implement this in vehicles and assets in the next phase.

Share this project:
×

Updates