Title: Innovative Strategies to Improve the Delivery of Telematics Voice Services
1Innovative Strategies to Improve the Delivery of
Telematics Voice Services
Julien Masson Director French Operations Connexis
LLC
2Consumers expectations for ubiquitous LBS
- Consumers are more more exposed to Navigation
systems - Boom of the PND market
- Imminent explosion of GPS phone segment
- Steady increase of OEM Navi systems
- They expect now to benefit from rich,
up-to-date, dynamic content ubiquitously with
easy-to-use location based services
3Many attempts from OEMs to provide easy-to-use
audio and visual in-vehicle services
- Using Navigation enhancement services whilst
driving require a convenient, easy-to-use, and
safe HMI (Audio, Visual, Tactile) - The Auto Industry has investigated over years
various paths to enable services - Embedded Voice Recognition
- Assistance through a live Operator
- More
4Remaining Cost and Quality Challenges of
In-Vehicle Audio Services
- Despite these attempts, a number of cost and
quality challenges remain for the delivery of
in-vehicle voice services - Embedded Voice Recognition
- Remains expensive with demanding CPU and storage
- Not yet mature to address Open Natural Language
- Assistance Through a Live Operator
- High Call Center operating cost vs. limited
customer value - No local knowledge from operators
5Key Industry Issue Expensive Call Centre costs
Expensive Call-Centre costs
Todays Call centre costs are expensive compared
to customers perceived value
Connexis innovation
- People-Assisted Computer Systems(PACS) reduce
traditional call centre costs, and streamlines
the customer experience
6PACS Flexible Call Center workflow
PACS is an innovative workflow processing
customers voice requests using a hybrid of
server-based speech recognition and human agents
in reserve to assist in interpreting unclear
commands
PACS Workflow
GPS Position / VIN
Request in Natural Language
- SRE reduces operational costs and
improves service delivery time - Silent agents backup SRE without
engaging the dialog with drivers - Customer satisfaction is never compromised
due to availability of L2 agent
Speech Recognition Engine (SRE)
Silent Agent (Level 1)
Request Process
Agent Conversation (Level 2)
7SRE Speech Shell (SS)Necessary data for SS
performance
- PACS Speech Shell requires three types of
database - Pronunciation data for POIs (dictionary)
- Acoustic speech data to create acoustic models
- Language data to create language models and
grammars
7
8PACS Return on Experience
-
- Nearly 1000 requests performed in French
language - 60 of requests are handled by Level1 and SRE
- SRE/L1 lt 50 sec., compared to 2 min. (industry
std.)
(Sec.)
Calls duration
Calls handling distribution
9Key Industry Issue No Local Knowledgeable agents
No Local Knowledgeable Agents
Todays Call centre agents are not the local
experts that drivers require
Connexis innovation
- Micro Call Centre model of small groups of
agents distributed geographically around regions
to enable the most local agent approach to
service delivery
10Distributed Service Delivery
- Network of agents located where customers drive
to provide the most locally knowledgeable service
experience
- Agents handle requests in native language with
local knowledge of roads, names, and landmarks - Local knowledge reduces call times and increases
customer satisfaction - High quality VoIP with EC/NR
- Sophisticated, high-speed, server-based call
routing, eliminating PBXs and ACDs - Ability to transfer calls smoothly
- 15 cost reduction vs. traditional call centre
11Applying PACS and Distributed Call Centres to
Telematics
- By combining speech recognition with human
agents, PACS - avoids the current dissatisfaction of man to
machine complex dialogs - reduces considerably labor costs
- Never compromises customer satisfaction
- Distributed call centers with work-at-home
agents promise to - lower facility costs
- solve the problems of scheduling and turnover
- improving service with local knowledge
Reductions in operation costs associated with
more efficient/accurate service delivery opens up
new business models