Autonomous (intelligent) software agents

Artificial Intelligence (AI) that incorporates Autonomous (Intelligent) Software Agents are being exploited more and more to resolve complicated challenges. The applicability of these solutions bring some benefits however, based on current focussed research trends and limitations associated, the opportunity and possibilities that this technology brings has not yet been fully realised. To look at this further we need to break down and understand the basic definition of what AI autonomous intelligent software agents are.

  • An AI autonomous intelligent software agents is anything that humans create that is capable of executing an activity based on information it perceives, (Carnegie Mellon University 2012) based on previous experiences and lastly is has to be autonomous to perform functions and make decisions how to execute the tasks it is asked to perform on its own.  (Brookshear, Smith and Brylow 2012)

AI                        (Terziyan) Some of the current trend areas of focus and applicability for AI autonomous intelligent software agents are:

  • Simulation of interaction’s and flows: Simulating environment usage, Film making, Digital animation or Video Gaming – Multi Agent Simulation System in Virtual Environment (MASSIVE) is the ability to use intelligent agents as crowds, oppositions, or movement simulations around buildings or roads. (MASSIVE Simulating Life 2011) such films as iRobot, Avatar have used (MASSIVE Simulaing Life 2013)  (Georgge 2009)
  • Interaction, Information retrieval, management and diagnosis: Agents that can communicate and create natural language speech responses, hypothesis generation, and evidence-based learning outcomes such as IBM Watson that is being used in healthcare diagnoses (Shacklett 2014)
  • E-Commerce platforms: Agents can be used in auction or buying sites such as amazon, ebay. Consumers can also use agents to shop for them, automated negotiations and decision-making. (Dignum et al. 2002)
  • Business process function management and planning: Planning or managing functions based on patterns of information such as scheduling on mass transportation systems. Scheduling or route planning of trains on the London Underground or analysis station patterns of movement. (Basra et al. 2005) (Skobelev 2011)

Some benefits of AI autonomous intelligent software agents based on the focus areas above can include:

  • Decentralisation of the tasks to agents allows each agent to execute tasks at their most reliable manner as they are able to. The agents are not tied to centralised system that will allocate and execute tasks as needed.
  • The capability to interact and connect to multiple systems and take feeds and react to information
  • Agents can become part of a large problem solving model by interacting with other agents

Some limitations of AI autonomous intelligent software agents:

  • Agents as stated are artificial creations that execute tasks assigned to them without doing anything greater than or less than the predefined task assigned.
  • Agents require input sensors to gather data, the more inputs the better the action can be defined and generate better outcomes. If inputs were limited the tasks can only be carried out based on simplest logic of the sensors available this could be limiting factor and could create errors. I.e. not creating enough complexity of the problem results in less complex answers. (Carnegie Mellon University 2012)

Example would be using IBM Watson as a doctor and only allowing it take inputs via speech. In this case if a patient had a chest infection and the patient said to IBM Watson “I have a temperature”, IBM Watson doesn’t have the ability to check the breathing of the patient’s chest that a regular doctor might have the result therefore from IBM Watson could be a wrong diagnosis to the patient “you have flu” instead of the a chest infection. (Shacklett 2014)

Developing technology such as Intelligent Software Agents isn’t easy and not that cheap to do, but over the past years this industry has grown significantly. Once they have been created though training an agent or creating groups of agents doesn’t need a very high level of skill. (MASSIVE Simulaing Life 2013) Therefore because of this companies have to see a business reason for using the technology to get a return on their investment. Software agents are in are actually more normal than we really think about.

  • Personal Assistants: Self-Contained systems that really look after that scheduling, organising and optimize people’s tasks. (Moraitakis 1997) (Salden 2013)
  • Visitor Hosting Systems: Carnegie Mellon University has a Visitor Hosting System. This system looks after events participated in by visitors at a site, and are organizes and coordinates through the cooperation of local agents with the visitors’ personal agents. (SYCARA and ZENG 1998)
  • Data Mining: This field is one of the fastest-evolving (as we read about Google and Facebook buying in to this technology) ones at the moment given the explosive growth of the amount of accessible information transmitted and received via networks and digital communications. (Bose 1998) (Yasemin 1999)
  • Network Management: Collaborative agents collect and exchange information on network statistics in order to achieve automation and optimization of network administration tasks these could be routing, access, service provisions, monitoring and statistical evaluation, within a global view or cloud platform. (Hermans 1996)
  • Air/Land Traffic Control: Multi-agent systems are installed in both controlling sites and travelling vehicles to help resolve control decisions in routing and scheduling. (Pinedo 1995)
  • Mobile technology: Agents to help with selecting and helping with finding frequently used functions or doing tasks (Mills and Stufflebeam 2005) (Intelligent and Mobile Agents Research Group)

Summary / Conclusions There are some outstanding use cases for AI autonomous intelligent software agents today and there is ongoing development and need to solve more and more complex challengers. This continuation and ongoing development of this technology in to new applications to support human lives, are being created all the time from mobile phones to films this also in turn sparks even more possibilities or ideas for this technology. The dream by some is to one day model human biology into a computing device thus no longer having a computing device be static logical thing but instead making a device that has “Intelligent and have the ability to think” and can even make decisions. One day this all might become reality and taking to a robot or system might become normal. (Brookshear, Smith and Brylow 2012)  


BASRA, Rajveer, et al. (2005). Resolving Schduling Issues of the London Underground using a Multi-agent system. [online]. Last accessed 24 05 2014 at:
BOSE, R. (1998). IDM: an intelligent software agent based data mining environment. [online]. Last accessed 27 05 2014 at:
BROOKSHEAR, J Gleen, SMITH, David T and BRYLOW, Dennis (2012). Chapter 11: Artificial Intelligence. In: HORTON, Marcia (ed.). Computer Science – An Overview. 11th ed., Addison-Wesley – Pearson, 461-500.
CARNEGIE MELLON UNIVERSITY (2012). Multi-Agent Systems. [online]. Last accessed 24 05 2014 at: DIGNUM,
Frank, et al. (2002). Software Agents Theory and Practice. [online]. Last accessed 24 05 2014 at: [online]. Last accessed 23 05 2014 at: GEORGGE, Matthew (2009). Massive What Is Massive. [online]. Last accessed 24 05 2014 at:
HERMANS, Björn (1996). Applications of Intelligent Agents. [online]. Last accessed 27 05 2014 at:
HERMANS, Björn (1997). Intelligent Software Agents on the Internet. [online]. Last accessed 24 05 2014 at: [online]. Last accessed 28 05 2014 at:
MASSIVE SIMULAING LIFE (2013). Films Gallery. [online]. Last accessed 24 05 2014 at:
MASSIVE SIMULATING LIFE (2011). Applications – What is Massive? [online]. Last accessed 24 05 2014 at:
MILLS, Frederick and STUFFLEBEAM, Robert (2005). Introduction to Intelligent Agents. [online]. Last accessed 28 05 2014 at:
MORAITAKIS, Nick (1997). Intelligent Software Agents Application and Classification. [online]. Last accessed 27 05 2014 at:
PINEDO, Michael (1995). OASIS (Optimal Aircraft Sequencing using Intelligent Scheduling). [online]. Last accessed 27 05 2014 at:
SALDEN, Alfons (2013). PERSONAL SOFTWARE AGENTS. [online]. Last accessed 27 05 2014 at:
SHACKLETT, Mary (2014). IBM Watson’s impressive healthcare analytics capabilities continue to evolve. [online]. Last accessed 24 05 2014 at:
SKOBELEV, Petr (2011). Bio-Inspired Multi-Agent Technology for Industrial Applications. [online]. Last accessed 24 05 2014 at:
SYCARA, KATIA and ZENG, DAJUN (1998). COORDINATION OF MULTIPLE INTELLIGENT SOFTWARE AGENTS. [online]. Last accessed 27 05 2014 at: [online]. Last accessed 24 05 2014 at:
YASEMIN, Ayse (1999). INTELLIGENT AGENTS: A DATA MINING PERSPECTIVE. [online]. Last accessed 27 05 2014 at:


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s